Smart Growth and Urban Goods Movement (2013)

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.

40 C H A P T E R 8 To gain a better understanding of the relationship between smart-growth land-use develop- ments and freight movements, the analytical tools available at the Puget Sound Regional Council (PSRC) were used to conduct a series of model runs. The tools available at PSRC include a blend of the state-of-the-art modeling tools, as well as more traditional analytical tools that would be familiar to other MPO or local jurisdictions. The tools were chosen because others could readily replicate the analysis in order to conduct policy analysis in their own regions. The modeling tools available at the PSRC are described in Appendix C, with specific attention paid to the tools used for the analysis discussed in this report. Currently PSRC is converting their forecasting tools from traditional four-step models to activity-based models. Trips are initially developed as tours but are treated as individual trips within the later modeling steps (destina- tion, mode, time of day, and route choice). Commercial vehicles are defined as any vehicle used for commercial purposes and can include autos, vans, sport utility vehicles, and small trucks, as well as medium and heavy trucks. These commercial vehicles are forecast using a truck model, which includes all commercial vehicles based on relative weight classes and separates light, medium, and heavy trucks for analysis purposes. This section begins with details regarding the specifications of the two land-use scenarios used for the analysis, as well as the three transportation-investment scenarios utilized. The section then describes the results of the modeling effort, with specific focus on the impacts of the smart- growth land-use scenario on truck miles of travel, truck hours of travel, truck delay, truck trip length and travel times, and emissions. The section concludes with comments on the implica- tions of these results for freight modeling and planning practice. 8.1 Description of the Land-Use Scenarios The land-use scenarios employed in this analysis represent two different policy outcomes for the year 2040, measured from a base year of 2000. As the land-use scenarios were developed to determine the impacts of population distribution resulting from two different long-range planning policies, an appropriately long time period was required to perform the analysis. The scenarios were developed as a part of the VISION 2040 planning process. VISION 2040 is a comprehensive growth-management, transportation, and economic development strategy developed for the Cen- tral Puget Sound Region, created to address the impacts of the sprawling development patterns that have emerged in the region. VISION 2040 centers around the Regional Growth Strategy, an expression of smart-growth policies, which outlines how various groupings of the region’s cities and unincorporated geographies should plan for additional population and employment growth. In particular, the growth strategy emphasizes growth in Regional Growth Centers and compact urban communities within the Urban Growth Area (VISION 2040 2012). Modeling

Modeling 41 Regional Growth Centers are a VISION concept, described as “locations characterized by compact, pedestrian-oriented development, with a mix of different office, commercial, civic, entertainment, and residential uses.”3 Centers are small geographies, generally centered around downtowns or other vibrant urban neighborhoods within cities. The State of Washington Growth Management Act requires cities and counties to designate Urban Growth Areas, which are intended to concentrate growth as a means of controlling urban sprawl, and the areas must have sufficient capacity for absorbing forecast growth. Figure 2 shows the distribution of Regional Growth Centers and the location of the Urban Growth Areas within the region. The scenarios for modeling these strategies apportion growth into a typology of cities and county areas, called Regional Geographies, which are defined as follows: • Metropolitan Cities—Metropolitan cities are central cities that serve as civic, cultural, and economic hubs in the region. They contain one or more Regional Growth Centers. • Core Suburban Cities—Core cities are other major cities that have Regional Growth Centers, serve as key multimodal transportation hubs, and already contain significant population and employment. They are intended to receive a significant share of future growth. • Larger Suburban Cities—Larger cities are larger inner-ring suburbs with a combined popula- tion and employment over 22,500 people and jobs. These cities contain important local and regional transit stations, ferry terminals, and other transportation connections. • Smaller Suburban Cities—Smaller cities represent a wide variety of communities from his- toric towns to growing new suburban cities, bedroom communities with limited employment activity and growth potential, and free-standing cities and towns separated from the region’s contiguous Urban Growth Area. • Unincorporated Urban Growth Area—This is land within the county’s jurisdiction within the designated Urban Growth Area. The unincorporated Urban Growth Area contains the largest amount of land area for any of the urban regional geography categories. These urban areas are quite diverse, with both lightly developed fringe areas and neighborhoods that are much more urban and nearly indistinguishable from surrounding incorporated jurisdictions. Approximately 60% of the land in the unincorporated Urban Growth Area has been affiliated with cities for future annexation. These areas, which are closely related to their adjacent city, are expected to accommodate a larger share of overall unincorporated urban growth than unaffiliated areas. • Rural Area—the Regional Growth Strategy calls for limiting the levels of development in the rural portions of counties, outside of the Urban Growth Area, in order to preserve rural char- acter and resource uses supported by rural levels of service and infrastructure. Two distinct development scenarios were created that attributed differing levels of develop- ment to the regional geographies to compare the impacts of broad policies such as smart growth and the Regional Growth Strategy on transportation investments. • Current Plans Extended—This is a “business-as-usual” scenario that extends current growth patterns, without changes, to 2040; this scenario relies on individual jurisdiction comprehen- sive plan targets. • Regional Growth Strategy—This is an expression of regional policy, countering past trends and refocusing growth in major cities and the densest urban areas, that is, regional smart growth. The differing levels of development apportioned to the regional geographies for the two strategies are shown in Table 5. The Current-Plans-Extended scenario extends land-use goals from the 2005–2025 compre- hensive planning cycle to 2040. Under the Growth Management Act (GMA), all cities in the 3 Puget Sound Regional Council. (2008). VISION 2040, p. 14.

42 Smart Growth and Urban Goods Movement Figure 2. Regional context map.

Modeling 43 central Puget Sound region must create comprehensive plans, which are to be updated every 7 years. The Washington State Office of Financial Management creates forecasts of population and employment growth for this time period, and cities craft comprehensive plan policies to accommodate this growth in accordance with policies within the GMA, as well as more specific regional and county planning policies. The Current-Plans-Extended scenario assumes that growth will continue in a fashion observed in the previous comprehensive planning cycle, before a true smart-growth policy scheme had been implemented. This planning cycle was guided by the previous version of VISION 2040 (VISION 2020), which expressed similar concepts of focusing growth into the Urban Growth Areas and Regional Growth Centers, but without as strong an architecture for distributing growth. Cities and counties would continue to encourage growth in urban centers around the region but also in unincorporated and rural areas, and to a wider degree than under the Regional Growth Strategy. Many of the region’s new jobs would locate in the largest cities, while medium-sized communities would also become larger employment centers. Many new apartments, condominiums, and town- houses would be built in downtown areas near job centers, and extensive residential growth would continue in the region’s unincorporated urban and rural areas. The distribution of population and employment for the Current-Plans-Extended scenario are shown Figure 3. The second scenario depicts the Regional Growth Strategy and represents an expression of smart-growth policies under which regional employment and housing growth are focused in cities that contain regionally designated growth centers. These regional centers are to be con- nected and served by a variety of transportation modes, including fast and frequent high-capacity transit service. Regional centers are intended to attract residents and businesses because of their proximity to services, jobs, well-designed housing, regional amenities, high-quality transit ser- vice, and other advantages. Locally designated town centers serve similar roles for smaller cities, but on a smaller scale than observed in the previous planning cycle (see Table 5). The Regional Growth Strategy also stresses preserving existing manufacturing and industrial centers. These are locations for intensive manufacturing, industrial, and related uses. Stressing employment growth in manufacturing and industrial centers, along with more active Regional Growth Centers and city centers, can help the region achieve a better jobs-housing balance, allow- ing people to live closer to their jobs, minimizing long commutes, lowering costs for extending new infrastructure, and limiting the effects of unbridled development on natural resources and rural lands. The distribution of population and employment are shown in Figure 4. 8.1.1 How the Scenarios Were Constructed 8.1.1.1 Current-Plans-Extended Scenario The distribution of the Current-Plans-Extended growth scenario was created by calculating the share that each city or unincorporated area had of the 2025 regional target total, 2025 being Scenario Metro Cies Core Cies Large Cies Small Cies Unincorporated UGA Rural Area Current Plans Extended—Populaon 26% 17% 9% 10% 24% 13% Current Plans Extended—Employment 45% 28% 7% 9% 8% 3% Regional Growth Strategy—Populaon 32% 22% 14% 8% 18% 7% Regional Growth Strategy— Employment 42% 29% 12% 6% 8% 2% Table 5. Percentage of 2040 regional growth by scenario.

Figure 3. Scenario one—Current Plans Extended.

Figure 4. Scenario two—Regional Growth Strategy.

46 Smart Growth and Urban Goods Movement the target horizon for the current planning cycle at the time of scenario creation. The assumption was that the share that a city or unincorporated area held in 2025 would remain fairly constant for the next 15 years, up to 2040. Growth was not allocated using each city or unincorporated area’s growth rate between 2000 and 2025 because of the severe recession that had occurred in the early 2000s. Separate methods were used to create the population and employment inputs. The general process for determining the land-use distributions is pictured in Figure 5. 8.1.1.1.1 Population. Step 1: Adjust base-year population. Three out of the four coun- ties used 2000 as the base year for setting their targets; only Snohomish County used 2002. To remain consistent among the counties, Snohomish’s 2002 base year had to be adjusted to 2000. The most viable option was to use Census 2000 population figures as a substitute for Snohomish County’s base. Step 2: Standardize population targets. Kitsap and Snohomish counties had targets set to 2025, while King and Pierce counties had targets out to 2022. The targets had to be adjusted so that each county’s numbers represented the year 2025. King and Pierce’s targets were straight- lined out from 2022 to 2025. To do this, first the change between the 2000 base and the 2022 targets was calculated (2022 target minus 2000 base). The change between 2000 and 2022 was then divided by the number of years from the base (22) to get the actual population growth per year, which was then multiplied by the number of years needed to get from 2022 to 2025 (i.e., 3). This number was then added to the original 2022 target to create the 2025 proxy target. Step 3: Determine city or unincorporated areas’ share of regional target total. Once all the target years were set to 2025, the regional target total was calculated by adding up the targets from the four counties. The share that each city or unincorporated area held of the regional target was then calculated by dividing the city or unincorporated areas’ targets by the regional population target total. For example, the regional target total for 2025 is 4,282,899, and Everett’s 2025 population target is 123,060, giving Everett roughly a 2.9% share of the regional population total (123,060/4,282,899). See Table 6 for regional percentage shares. Step 4: Distribute regional forecast change from 2025 to 2040. Using the calculated population share for each city or unincorporated area, the change between the 2025 regional population target total and the 2040 regional forecasted population total (705,101) was distributed. Using Everett again as an example, the 2.9% share that Everett held gave it approximately 20,260 of the regional population change. Adding this to Everett’s 2025 target (123,060) gives the city a 2040 population total of 143,320. See Table 6 for 2040 population totals. 8.1.1.1.2 Employment. Step 1: Standardize base-year employment. Only two of the four counties, King and Snohomish, had set employment targets. Kitsap and Pierce did not have 2000 base-year employment numbers. To create a standardized base year, the 2000 employment data were used as the base for all four counties. Step 2: Create proxy targets for Kitsap and Pierce Counties. To create proxy employment targets for Kitsap and Pierce, a straight-line method was used to produce a 2020 target. The 1995–2004 average annual change for employment was calculated. Because of significant fluctuations in the economy during those years, the city or unincorporated areas’ shares of the county total from the Standardize base year 2000 Standardize regional targets to 2025 Determine city share of 2025 regional target Apply share to 2025- 2040 change Add to the 2025 city target Figure 5. Process for creating the Current-Plans-Extended scenario.

Jurisdicon % of Regional 2040 Populaon % of Regional 2040 Employment Jurisdicon % of Regional Pop. Target 2040 Populaon % of Regional 2040 Employment KING COUNTY 48.90% 2,440,420 66.60% 2,045,207 PIERCE COUNTY 22.00% 1,096,635 15.10% 463,637 Unincorporated 9.10% 455,639 2.40% 73,479 Unincorporated 9.30% 465,994 3.40% 105,634 Uninc Rural 3.50% 173,001 0.80% 23,370 Uninc Rural 3.80% 188,075 0.80% 25,861 Uninc UGA 5.70% 282,638 1.60% 50,109 Uninc UGA 5.60% 277,919 2.60% 79,773 Incorporated 39.80% 1,984,782 64.20% 1,971,728 Incorporated 12.60% 630,641 11.70% 358,002 Algona 0.10% 3,831 0.10% 2,816 Auburn (Prc) 0.20% 10,497 0.00% 627 Auburn (Kin) 1.30% 64,103 2.10% 63,522 Bonney Lake 0.50% 23,194 0.20% 6,516 Beaux Arts 0.00% 358 0.00% 25 Buckley 0.10% 6,224 0.10% 4,502 Bellevue 3.08% 153,675 7.40% 226,399 Carbonado 0.00% 1,000 0.00% 119 Black Diamond 0.20% 8,223 0.10% 3,938 Du Pont 0.20% 11,654 0.20% 7,223 Bothell (Kin) 0.50% 23,786 0.60% 17,556 Eatonville 0.10% 3,360 0.10% 1,546 Burien 0.80% 40,372 0.60% 19,822 Edgewood 0.30% 16,688 0.10% 2,957 Carna€on 0.10% 3,180 0.00% 1,136 Fife 0.20% 11,019 0.70% 20,952 Clyde Hill 0.10% 3,366 0.00% 613 Fircrest 0.20% 8,068 0.10% 2,170 Covington 0.40% 20,803 0.20% 4,751 Gig Harbor 0.30% 13,265 0.50% 14,301 Des Moines 0.80% 38,320 0.30% 10,684 Lakewood 1.70% 86,043 1.40% 42,036 Duvall 0.20% 9,112 0.10% 2,801 Milton (Prc) 0.20% 8,473 0.10% 3,624 Enumclaw 0.40% 19,501 0.30% 7,872 Or…ng 0.20% 9,858 0.10% 2,319 Federal Way 2.30% 113,668 1.70% 52,153 Pacific (Prc) 0.00% 0 0.10% 4,283 Hunts Point 0.00% 516 0.00% 51 Puyallup 0.90% 45,842 1.30% 38,890 Issaquah 0.50% 24,267 1.30% 39,539 Roy 0.00% 1,211 0.00% 392 Kenmore 0.60% 28,436 0.30% 10,040 Ruston 0.00% 2,212 0.00% 446 Kent 2.10% 103,032 3.30% 101,785 South Prairie 0.00% 1,038 0.00% 274 Kirkland 1.30% 65,626 2.20% 67,727 Steilacoom 0.20% 8,171 0.10% 2,848 Lake Forest Park 0.30% 16,642 0.10% 2,548 Sumner 0.30% 14,862 0.50% 16,254 Maple Valley 0.30% 16,972 0.20% 5,022 Tacoma 6.20% 307,056 5.70% 175,983 Medina 0.10% 3,507 0.00% 516 University Place 0.80% 40,243 0.30% 9,649 Mercer Island 0.60% 29,581 0.30% 10,712 Wilkeson 0.00% 665 0.00% 91 Milton (Kin) 0.00% 1,054 0.00% 1,390 SNOHOMISH 21.30% 1,064,763 13.50% 416,261 Newcastle 0.20% 11,472 0.10% 2,131 Unincorporated 11.60% 578,503 2.90% 88,076 Normandy Park 0.10% 7,444 0.00% 910 Uninc Rural 4.40% 218,410 0.70% 20,923 North Bend 0.20% 10,200 0.20% 4,777 Uninc UGA 7.20% 360,093 2.20% 67,153 Table 6. Current Plans Extended population totals and shares. (continued on next page)

Pacific (Kin) 0.20% 8,706 0.00% 1,422 Incorporated 9.70% 486,261 10.70% 328,185 Redmond 1.60% 77,984 4.10% 126,538 Arlington 0.40% 20,218 0.60% 17,480 Renton 1.50% 74,081 3.80% 115,970 Bothell (Sno) 0.50% 25,622 0.60% 19,291 Sammamish 1.10% 53,046 0.30% 8,497 Brier 0.20% 9,072 0.00% 537 SeaTac 0.90% 43,484 1.90% 57,280 Darrington 0.00% 2,224 0.00% 560 Seaƒle 15.70% 783,068 28.00% 859,022 Edmonds 1.00% 52,269 0.50% 15,505 Shoreline 1.40% 68,345 0.80% 24,838 Evere 2.90% 143,320 5.10% 158,109 Skykomish 0.00% 289 0.00% 153 Gold Bar 0.10% 3,374 0.00% 260 Jurisdicon % of Regional 2040 Populaon % of Regional 2040 Employment Jurisdicon % of Regional Pop. Target 2040 Populaon % of Regional 2040 Employment Snoqualmie 0.20% 8,042 0.10% 4,403 Granite Falls 0.10% 5,555 0.10% 2,416 Tukwila 0.60% 28,943 2.90% 90,207 Index 0.00% 221 0.00% 86 Woodinville 0.30% 16,531 0.70% 22,093 Lake Stevens 0.20% 9,736 0.10% 2,157 Yarrow Point 0.00% 1,214 0.00% 71 Lynnwood 0.90% 44,850 1.30% 39,337 KITSAP COUNTY 7.70% 386,181 4.80% 147,096 Marysville 0.90% 46,259 0.50% 15,122 Unincorporated 4.80% 239,604 1.70% 53,204 Mill Creek 0.40% 18,738 0.20% 5,507 Uninc Rural 2.90% 142,478 1.10% 34,905 Monroe 0.50% 23,922 0.50% 14,283 Uninc UGA 1.90% 97,127 0.60% 18,299 Mountlake 0.50% 26,153 0.30% 10,309 Incorporated 2.90% 146,577 3.10% 93,892 Mukilteo 0.50% 25,622 0.40% 11,699 Bainbridge Island 0.70% 33,378 0.40% 12,005 Snohomish 0.20% 11,624 0.20% 6,228 Bremerton 1.20% 60,581 1.60% 48,941 Stanwood 0.10% 6,580 0.20% 5,702 Port Orchard 0.30% 13,152 0.30% 8,174 Sultan 0.20% 9,538 0.10% 3,484 Poulsbo 0.20% 12,289 0.30% 9,345 Woodway 0.00% 1,363 0.00% 113 Silverdale 0.50% 27,177 0.50% 15,426 Table 6. (Continued).

Modeling 49 2004 employment data were also calculated. This percentage was then averaged with the aver- age annual change between 1995 and 2004 to lessen the significant swings in the economy. The 2020 small-area forecast totals for each county were then distributed by the calculated change. For example, between 1995 and 2004, Bremerton’s average annual change in employment was 30.3%. In 2004, the city accounted for 36.2% of Kitsap County’s total employment. Averaging these two percentages together gives a percentage growth of 33.3. Kitsap County’s year 2020 small-area fore- cast is 116,865, and Bremerton’s share is 33.3% (33.27154%) of that, or 38,883. Step 3: Roll back King and Snohomish Counties’ targets to 2020. To achieve consistent employ- ment targets, King and Snohomish County targets were rolled back to 2020 and based on the small-area forecasts, as were Kitsap and Pierce Counties’. To achieve this, the annual change of each county’s targets was calculated, and this change was then multiplied by the number of years the target needed to be rolled back (King, 2 years; Snohomish, 5 years) to determine the amount of employment to be subtracted from the target. Once the employment was subtracted, the 2020 targets had to be controlled to the county small-area forecast totals. The percentage share of the city or unincorporated area of the county’s total 2020 target was calculated. For example, the city of Everett had a 38% share of the county’s total employment. (Everett = 120,495; Snohomish County = 317,233). The 2020 county small-area forecast total was then allocated based on these percentage shares. Step 4: Determine city or unincorporated areas’ shares of regional target total. Once all the tar- get years were set to 2020, the regional target total was calculated by adding up the targets from the four counties. The share that each city or unincorporated area held of the regional target was then calculated by dividing the city or unincorporated areas’ targets by the regional employ- ment target total. For example, the regional target total for 2020 is 2,278,603, and Everett’s 2020 employment target is 117,267, giving Everett roughly a 5.1% share of the regional employment total (117,267/2,278,603). See Table 6 for regional percentage shares. Step 5: Distribute regional forecast change from 2020 to 2040. Using the calculated employ- ment share for each city or unincorporated area, the change between the 2020 regional employ- ment target total and the 2040 regional forecasted employment total (793,597) was distributed. Using Everett again as an example, the 5.1% share that Everett held gave it approximately 40,842 of the regional employment change. Adding this to Everett’s 2020 target (117,267) gives the city a 2040 employment total of 158,109. See Table 6 for 2040 employment totals. 8.1.1.2 Regional Growth Strategy Distributions of population and employment in the Regional Growth Strategy alternative were originally created using the sketch planning tool Index. The region was divided into 150-square- meter grid cells, to which users then ascribed one of 26 defined land uses, each of which carried specific population and employment values. Land uses were applied to demonstrate a particular distribution of population and employment to the region’s cities and counties in the year 2040. Once the land-use inputs were distributed to the grid cells, these data were aggregated into Trans- portation Analysis Zones (TAZs) and Forecast Analysis Zones (FAZs) to be used as inputs to the region’s Transportation Demand and Air Quality models. In addition, characteristic data required by the travel model on group quarters, income quartiles, and employment sectors were also added. The following procedure was used to develop the detailed data inputs from the Index base data. • Convert the Index base geography from grid cells and cities to FAZs. PSRC’s Small-Area Forecast model is zone based and structurally limited to the 219 zones that the Puget Sound Region is currently divided into. The first step, then, was to sum the Index 2000 base-year data and the 2040 future-year data for each scenario for each of the 219 FAZs. • Expand the Index 2000 base-year data into the detailed data variables. The original year 2000 data used for Index contained characteristic data on employment sector, household size, and

50 Smart Growth and Urban Goods Movement income required to run the model, and the characteristic data were translated to percentage shares. For example, in year 2000, the percentage of the jobs in each FAZ that were Retail, Manufacturing, and so forth, were estimated. • Apply the growth projected in the 2003 Small-Area Forecasts to each of the characteristic data variables. The PSRC Small-Area Forecasts from 2003 include 2000 and 2030 forecasts by FAZ for each of the characteristic data variables. Using these endpoints, the percentage share that each characteristic variable represented of the overall base-data category was determined, extrapolated to the year 2040, and applied to the year 2040 Index base category total for that FAZ. For example, if the Small-Area Forecasts showed the percentage of Low-Income- Quartile households in FAZ 110 dropping from 25% to 18%, that reduction in “share size” was then applied to the Index data in 2040 to arrive at a preliminary estimate. • Balance the preliminary estimates with the regional forecasts for 2040. PSRC’s forecast process is top-down, with the regional demographic and economic forecasts determined first and then allocated to a sub-regional geography. To control based on these forecasts, a factoring process adjusted the Index-based 2040 FAZ-level detailed data so that each Index scenario, as modeled, would match the regional forecasts as well. 8.2 Travel Network Scenarios Three discrete transportation networks were modeled to accompany the two land-use sce- narios. The transportation networks were intended to capture a status quo scenario, one that favors smart-growth investments, and one that favors traditional roadway investments. The initial intent of the modeling exercise was to determine the impacts of land use on freight transportation. However, because smart-growth principles include transportation efficiency and land-use design, a transportation scenario reflecting a smart-growth orientation was included to model the interaction between land use and transportation. Finally, for the sake of completeness, a highway-heavy transportation-investment scenario was also evaluated. The transportation networks were developed originally for Transportation 2040, the Puget Sound region’s long-range transportation plan adopted in 2010, and they were vetted as part of that pro- cess. The baseline scenario reflected existing conditions in the region and was chosen as the status quo alternative. Alternative Two of the original scenario analysis used for plan development was employed to characterize the roadway-heavy investment scenario type. The final, adopted, preferred alternative from the planning work was chosen as the smart-growth scenario because of its reliance on tolling and transit investments. These three scenarios are described in greater detail below. 4 8.2.1 Baseline Alternative (Status Quo) The baseline transportation network consists of the existing transportation systems and a limited series of future investments. This alternative is meant to illustrate what would most likely occur with the transportation system assuming there were no interventions. It is a useful starting point for comparison with the other transportation alternatives and is also a constant that allows for comparisons across different land-use policies (e.g., status quo vs. smart growth). The baseline Alternative is funded almost completely with “current law” traditional revenue sources—gas tax, sales tax, state and federal grants and loans, local general fund revenues, permit and licensing fees, and limited tolling (on the Tacoma Narrows Bridge and the auto ferry 4 A complete discussion of the alternatives developed through the Transportation 2040 planning process and the specific facili- ties and investments included in each alternative may be found in Appendix A of the Final Environmental Impact Statement of the plan: http://psrc.org/assets/3694/Appendix_A_-_Transportation_2040_Alternatives_Report.pdf.

Modeling 51 system). The baseline Alternative would build state highway projects funded under the state’s “nickel” gas tax and Transportation Partnership Account (TPA) programs, plus Sound Transit’s Phase 2 plan (ST2), approved by voters in November 2008. It would sustain existing ferry service and demand-management programs and make modest additions to transit service, including King County Metro’s Rapid Ride and Community Transit’s Swift bus rapid transit (BRT) sys- tem. Beyond “current law” funding, the baseline Alternative assumes that the region would find sufficient additional revenue to fully maintain and preserve the existing transportation system. 8.2.2 Alternative Two (Roadway Investments) Of all the plan alternatives, Alternative Two adds the most roadway capacity through lane additions to existing highways, the creation of several new highways (SR 167 Extension, SR 509 Extension, and the Cross-Base Highway), and additional lanes on the regional arterial network. It adds considerable light rail capacity and a new auto ferry route across Puget Sound. It adds pedestrian and bicycle infrastructure in key locations. Its demand-management, bus service, and system-management investments are similar to the baseline Alternative. Its most significant management strategy is the establishment of a two-lane High Occupancy Toll (HOT) system on much of the regional freeway network (with some one-lane HOT facilities) to manage con- gestion and provide revenue to supplement traditional funding sources. Traditional funding sources would provide the majority of the financing. 8.2.3 Preferred Alternative (Smart Growth) The Preferred Alternative is designed to improve the region’s transportation system through a combination of investments in system efficiency; strategic expansion; transit, ferry, bike, and pedestrian improvements; and investments to preserve the existing transportation system. The Preferred Alternative financial strategy is based on a phased approach of transitioning away from current gas taxes toward the implementation of new user fees. The Preferred Alternative includes the following: • More transit service than all other alternatives • More miles of biking and walking facilities focused on access to transit stations and centers and completion of regional trail links than all other alternatives • Current levels of vehicle ferry service and additional passenger ferries • Replacement of several vulnerable roadways, including the Alaskan Way Viaduct and the SR 520 Floating Bridge • Completion of missing links in the highway network such as SR 509, SR 167, and the Cross- Base Highway • Expansion of local arterials and state highways in limited but strategic ways to service growth in Urban Growth Centers 8.3 Modeling Results Six model runs were conducted to provide better understanding of the relationship between smart-growth land use, transportation system investments, and truck travel. The major findings, which are described in further detail throughout the remainder of the section include the following: • Truck miles of travel are consistently lower under the smart-growth land-use scenario than under the alternative (status quo), regardless of time of day, roadway facility type, truck type, or transportation network investments. • Truck hours of travel are consistently lower under the smart-growth land-use scenario than under the alternative, regardless of time of day, roadway facility type, truck type, or

52 Smart Growth and Urban Goods Movement transportation-network investments. The lowest daily truck hours of travel appear in the scenario that combines the smart-growth land use with the smart-growth transportation investments (transit and non-motorized). • Truck delay is seen to be somewhat higher in the smart-growth land-use scenario, but improve- ments to the transportation network have a pronounced impact in terms of delay reductions, with the smart-growth improvements performing better than the roadway investments. • Truck trip lengths remain relatively flat across origin and destination pairs involving the movement of goods. However, when the destination involves a concentrated smart-growth area, truck trip lengths are longer under the smart-growth land-use scenario. • Truck travel times remain relatively flat across origin and destination pairs involving the movement of goods. However, when the destination involves a concentrated smart-growth area, truck travel times are longer under the smart-growth land-use scenario. • Pollution emission levels, particularly for carbon dioxide, are consistently lower in the smart- growth land-use scenario as compared with the baseline land-use scenario. Emission levels are at the lowest point when the smart-growth land-use scenario is coupled with smart-growth transportation investments. 8.3.1 Truck Miles of Travel Across all three transportation networks, the smart-growth land-use patterns produce lower daily truck miles of travel (see Figure 6). This trend is also consistent across individual time periods, not just across an entire day (see Table 7). 12 12.5 13 13.5 14 14.5 Baseline Road Investments Smart Growth Transportation Network Da ily T ru ck M ile s o f T ra ve l ( M ill io ns ) Baseline Land Use Smart Growth Land Use Figure 6. Daily truck miles of travel. Transportation Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth a.m. 2.7 2.6 2.9 2.8 2.9 2.8 Midday 5.6 5.4 5.9 5.7 5.9 5.8 p.m. 2.7 2.6 2.9 2.8 2.9 2.9 Evening 1.3 1.2 1.3 1.3 1.3 1.3 Night 1.0 1.0 1.1 1.0 1.1 1.0 Table 7. Daily truck miles of travel (millions) by time period.

Modeling 53 Notably, truck miles of travel are higher in the altered transportation networks (roadway and smart-growth investments) as compared with the baseline transportation network. These results may appear counterintuitive and presumably are due to the improvements to the overall trans- portation network leading to a rise in demand for travel. Both altered transportation networks see a marked increase in home-based work trip generation and distribution. The availability of improved roadway facilities stimulates the use of those facilities, and improved transit and non- motorized transportation facilities reduce passenger vehicle miles of travel. If a policy goal were to reduce truck miles of travel, the conclusion should not be drawn that investment in transportation improvements increases truck travel. In all three scenarios, truck miles of travel are lower under the smart-growth land-use scenario as compared with the alternative. In addition to the time of day, truck miles of travel are consistently lower across transporta- tion facility type for the smart-growth land-use scenario than the alternative (see Table 8). Although the truck miles of travel are consistently lower under the smart-growth land-use scenario as compared with the alternative, freeway travel increases and arterial travel decreases under the two improved transportation networks. For the investments in roadway facilities, the improved freeway facilities provide more favorable, less congested, and faster routes than were previously available. The smart-growth transportation investments stimulate a mode shift away from SOVs and again open up capacity on the freeways. However, truck travel on the connec- tor facilities (smaller local roads) remains unchanged across all of the transportation invest- ments. This is most likely because (1) truck origins and destinations are fixed and must use local facilities to arrive at the arterial and freeway facilities, and (2) certain types of trucking activities (e.g., package delivery, waste management) must travel on all roads for their freight-hauling purposes, creating an inelastic demand for use of those facilities. The same trends in truck miles of travel by time of day and facility type are also seen across different types of trucks, not just the aggregate daily total (see Table 9). For each truck type (light, medium, or heavy) overall miles of travel are uniformly lower under the smart-growth land-use scenario than under the alternative, and this is also the case by time of day and facility Transportaon Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Freeway 9,272 8,908 10,567 10,228 10,672 10,404 Arterial 3,066 3,019 2,608 2,534 2,550 2,462 Connector 974 934 976 936 976 936 Total 13,312 12,861 14,150 13,698 14,199 13,802 Table 8. Daily truck miles of travel (thousands) by facility type. Transportaon Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Light 3,090 2,931 3,656 3,509 3,685 3,587 Medium 4,950 4,772 5,099 4,904 5,140 4,949 Heavy 5,272 5,158 5,395 5,284 5,373 5,267 Table 9. Daily truck miles of travel (thousands) by truck type.

54 Smart Growth and Urban Goods Movement type. Transportation investment (roadway or smart growth) shifts more of each truck type to freeways and away from arterials, while each truck type’s miles of travel remains constant on collector facilities. 8.3.2 Truck Hours of Travel Similar to truck miles of travel, total daily truck hours of travel are consistently lower (although only by a small amount) in the smart-growth land-use scenario than the alternative (see Figure 7). However, unlike truck miles of travel, investments in the transportation system considerably reduce overall truck hours of travel. This result is likely due to improved capacity of the transportation facilities, especially due to shifts away from non-single-occupancy travel for passenger modes because the smart-growth investments (transit and non-motorized) have a much more pronounced effect than the roadway-capacity improvements. For the most part, the same results seen for total daily hours of truck travel can be seen across the various time periods—the smart-growth land-use scenario performs better than the 0 50 100 150 200 250 300 350 400 450 500 Baseline Road Investments Smart Growth Transportation Network Th ou sa nd s Baseline Land Use Smart Growth Land Use Figure 7. Daily truck hours of travel. Transportaon Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth a.m. 98,174 96,500 88,400 86,311 77,847 76,821 Midday 187,734 195,332 177,634 170,024 154,844 150,975 p.m. 113,721 105,466 102,819 103,555 85,604 86,094 Evening 37,768 35,273 36,830 35,759 34,513 33,927 Night 29,490 28,112 27,813 28,479 26,690 26,449 Bold indicates instances in which the baseline land-use scenario performs be­er than the smart-growth land-use scenario. Table 10. Daily truck hours of travel by time period.

Modeling 55 alternative (see Table 10). However, there are several exceptions (shown in bold) in which the smart-growth land-use scenario has a slight increase in truck hours of travel over the status quo scenario. The difference in hours of travel between the land-use scenarios is relatively small across the time period, but significant reductions are achieved by the transportation invest- ments, particularly the smart-growth ones (transit and non-motorized) over roadway invest- ments and no transportation investments. In terms of truck performance, the fewest truck hours of travel are seen under the smart-growth land-use scenario with commensurate smart-growth investments in the transportation system. Truck hours of travel on different transportation facilities are also consistently lower under the smart-growth land-use scenario as compared with the alternative (see Table 11). Unlike miles of travel, the decrease in hours of travel is uniform across facilities and investments. Under the roadway and smart-growth transportation investments, both freeway and arte- rial hours of travel are reduced. And again, the biggest impact in terms of reduction of truck hours of travel is present under a smart-growth land-use scenario paired with transit and non-motorized transportation investments. Similar to truck miles of travel, hours of travel on collector streets are unchanged across transportation scenarios, reflecting the inelastic demand for those facilities. Truck hours of travel for individual truck classes show similar results (see Table 12). The smart-growth land-use scenario generally provides a reduction in truck hours of travel from the level of the baseline land-use scenario. In addition, the smart-growth land-use scenario coupled with investments in transit and non-motorized transportation improvements leads to the largest potential reductions in overall truck hours of travel for all three classes of trucks as compared with the other potential alternatives. Transportaon Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Freeway 270,881 267,273 260,905 254,958 211,676 210,489 Arterial 146,579 146,622 123,049 122,279 118,238 116,849 Connector 49,427 46,789 49,543 46,890 49,586 46,928 Total 466,887 460,683 433,496 424,128 379,499 374,265 Bold indicates instances in which the baseline land-use scenario performs beer than the smart-growth land-use scenario. Table 11. Daily truck hours of travel by facility type. Transportaon Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Light 124,244 120,097 124,155 121,741 114,102 112,911 Medium 178,786 176,064 160,643 156,145 137,224 134,091 Heavy 163,858 164,523 148,698 146,242 128,173 127,263 Bold indicates instances in which the baseline land-use scenario performs be er than the smart-growth land-use scenario. Table 12. Truck hours of travel by truck type.

56 Smart Growth and Urban Goods Movement 8.3.3 Truck Delay Overall daily delay for trucks is slightly higher for the smart-growth land-use scenario than under the baseline scenario (see Figure 8). Across all three sets of transportation systems (base- line, road, and smart-growth transportation investments) the delay under the smart-growth land-use scenario is roughly one half of 1% higher than the baseline land use. Given the mag- nitude of overall system delay, the results in delay between the two land-use scenarios is essen- tially indistinguishable. However, investment in the transportation system has a striking effect on delay, with the roadway investments reducing daily delay by 24% over the baseline, and transit and non-motorized investments reducing delay by 54% over the baseline transporta- tion scenario. These results are repeated for the freeway and arterial facilities, in addition to the overall network. The performance of the smart-growth (i.e., the Preferred) land-use scenario as compared with the baseline scenario in terms of delay has a fair amount of variance across time periods and transportation investments (see Table 13). There does not appear to be a distinguishable pattern across the transportation-investment scenarios in which specific time periods have more delay 0 20 40 60 80 100 120 140 160 180 200 Baseline Road Investments Smart Growth Transportation Network Th ou sa nd s Baseline Land Use Smart Growth Land Use Figure 8. Total daily truck delay (hours). Transportaon Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth a.m. 39,604 39,732 27,995 27,911 17,172 17,962 Midday 69,320 80,369 54,972 51,458 32,162 31,962 p.m. 53,522 47,496 40,620 43,252 23,243 25,551 Evening 10,677 9,255 8,450 8,370 6,136 6,418 Night 7,702 7,133 5,048 6,399 4,056 4,478 Bold indicates instances in which the baseline land-use scenario performs be­er than the smart-growth land-use scenario. Table 13. Daily delay by time period.

Modeling 57 in one land-use scenario over the other. However, for the smart-growth transportation invest- ments (transit and non-motorized), the smart-growth land-use scenario seems to have slightly more delay than the baseline scenario in most time periods despite the small overall impact. The exception to the delay results is seen when examining delay by truck type (see Table 14). Medium and heavy trucks perform slightly better under the smart-growth land-use scenario compared with the baseline scenario in the context of the roadway-investment transportation scenario. This result follows the logic that an investment in freeway facilities will improve condi- tions for all users of the transportation system, including freight users. 8.3.4 Truck Trip Lengths To understand the impact of the modeled land-use and transportation scenarios on truck trip length, the transportation analysis zones used in the analysis were separated into four categories as follows: 1. Smart-growth zones have high densities and are balanced in terms of jobs and housing. 2. Goods-dependent zones have a high concentration of freight-related employment (ware- housing, communication, transportation, utilities). 3. The most concentrated goods-dependent zones have the highest concentration of freight- related employment. 4. All other zones were parts of the region that had neither significant freight-related employment nor other concentrations of activities. The typology of analysis zones allowed consideration of the truck trip length by types of trips. For example, manufacturing trips or drayage trips likely would occur between two sets of goods-dependent locations. For trips originating and terminating within the region’s most con- centrated freight-related analysis zones, average trip lengths remained almost constant across the two land-use scenarios and the three different transportation-investment scenarios (see Table 15). In contrast, trips originating in the region’s most concentrated centers of freight activity and terminating in the smart-growth locations (downtown cores, urban villages, etc.), uniformly have longer distances under the smart-growth land-use scenario as compared with the baseline in all three of the transportation-investment scenarios (see Table 16). The preceding discus- sion of truck miles of travel showed that the smart-growth scenario consistently demonstrated fewer overall miles. This further examination of trip origin and destination highlights the results that, although overall trip length and truck miles of travel are reduced under the smart-growth land-use scenario, specific trips (which are likely to be deliveries to urban cores) would have longer trip distances under this type of land-use configuration. Indeed, trips between the most concentrated areas and the less-goods-dependent areas, as well as those analysis zones that are Transportaon Baseline Roadway Investments Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Light 45,844 45,277 37,468 38,363 27,201 28,567 Medium 73,818 74,714 54,593 54,178 30,928 31,828 Heavy 61,163 63,994 45,023 44,850 24,639 25,975 Bold indicates instances in which the baseline land-use scenario performs be er than the smart-growth land-use scenario. Table 14. Daily delay by truck type.

58 Smart Growth and Urban Goods Movement not centers of activity, all have shorter truck trip lengths across all the time periods under the smart-growth land-use scenario. 8.3.5 Travel Times In most cases, only minor differences in average travel times between different types of zones are observed when comparing the baseline and smart-growth land-use scenarios (see Table 17). The most notable change in travel times appears to be due to transportation investments rather than the configuration of land uses. As has been the case elsewhere, investments in roadway facilities reduce travel times, but investments in transit and motorized modes of transportation reduce travel times even further. The most pronounced instances in which the smart-growth land-use scenario travel times exceed those of the baseline land-use scenario occur when the zonal pairs involve origins related to goods movement and the destination zones are Smart-Growth Centers, for example, down- town cores and urban villages (see Figure 9). This result is indicative of the great demand placed on the transportation network around smart-growth areas, where trucks must compete with all other vehicles for right of way, thus creating longer travel times to these locations. Transportaon Investment Baseline Roadway Smart Growth From To Time Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Most Concentrated Most Concentrated M 23.6 23.5 23.7 23.7 23.8 24.0 Most Concentrated Most Concentrated MD 23.6 23.6 23.7 23.7 23.7 23.8 Most Concentrated Most Concentrated PM 23.7 23.5 23.8 23.7 24.1 24.2 Most Concentrated Most Concentrated EV 23.7 23.7 23.9 23.9 23.4 23.4 Most Concentrated Most Concentrated NT 23.6 23.7 23.9 23.7 23.6 23.6 Bold indicates instances in which the baseline land-use scenario performs beer than the smart-growth land-use scenario. Table 15. Average freight-related truck trip lengths. Transportaon Investment Baseline Roadway Smart Growth From To Time Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Most Concentrated Smart Growth AM 23.4 25.4 23.9 25.9 23.2 25.5 Most Concentrated Smart Growth MD 22.9 25.1 23.9 26.0 23.0 25.0 Most Concentrated Smart Growth PM 22.8 25.0 23.5 25.0 22.9 24.9 Most Concentrated Smart Growth EV 23.8 26.0 24.0 26.1 23.4 25.6 Most Concentrated Smart Growth NT 23.7 25.9 24.0 25.9 23.7 25.8 Bold indicates instances in which the baseline land-use scenario performs be­er than the smart-growth land-use scenario. Table 16. Average smart-growth-related truck trip lengths.

Modeling 59 Transportaon Investment Baseline Roadway Smart Growth From To Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth Most Concentrated Most Concentrated 50.0 50.1 42.2 43.2 36.7 37.3 Most Concentrated Goods Dependent 61.8 60.6 52.7 53.0 46.1 46.1 Most Concentrated Smart Growth 50.9 54.5 43.9 47.6 38.0 41.1 Most Concentrated All Else 66.4 66.5 57.3 58.6 50.5 51.6 Goods Dependent Goods Dependent 70.5 69.0 61.0 60.9 53.4 53.1 Goods Dependent Most Concentrated 62.0 61.3 53.2 53.5 46.3 46.3 Goods Dependent Smart Growth 61.6 64.5 53.8 56.8 46.5 49.2 Goods Dependent All Other 74.8 74.4 65.3 66.1 57.6 58.3 Bold indicates instances in which the baseline land-use scenario performs beer than the smart-growth land-use scenario. Table 17. Average truck travel times (minutes). 0 10 20 30 40 50 60 70 Baseline / Baseline Smart Growth / Baseline Baseline / Roadway Smart Growth / Roadway Baseline/ Smart Growth Smart Growth / Smart Growth Most Concentrated to Most Concentrated Most Concentrated to Goods Dependent Most Concentrated to Smart Growth Most Concentrated to All Else Land-Use Scenario/Transporta on Investment M in ut es Figure 9. Average zonal truck travel times. 8.3.6 Air-Quality Benefits Truck miles and hours of travel and truck delay have immediate impacts on trucking firms and goods-dependent businesses. A reduction in these costs ultimately has secondary impacts related to the price of goods and the economy. Similar to those secondary benefits, improve- ments in air quality present a societal benefit worth considering through the lens of the various land-use and transportation-investment scenarios.

60 Smart Growth and Urban Goods Movement The smart-growth land-use scenario creates large benefits in terms of emissions reductions as compared with the baseline land-use scenario. Table 18 summarizes the reduction from all the scenarios in terms of tons reduced (or gained) as compared with the baseline land-use scenario and the baseline transportation network. Not surprisingly, the roadway-investment transportation scenario incentivizes further driving, as evidenced by increased vehicle and truck miles of travel, which leads to an increase in most of the reported pollutants. However, because of the efficiency gains, as seen through reduced truck hours of travel, carbon dioxide emissions are decreased even under the roadway-investment scheme. Regardless, even under the roadway-investment transpor- tation scenario, as is the case with the other transportation scenarios, emissions are consistently reduced more under the smart-growth land-use scenario than under the alternative. The most notable reduction in pollution comes in the form of carbon dioxide (see Fig- ure 10). The smart-growth land-use scenario has greater emission reductions under all three Transportation Baseline Roadway Smart Growth Land Use Baseline Smart Growth Baseline Smart Growth Baseline Smart Growth CO 1,294.09 (19.69) 86.51 72.12 (14.74) (24.80) NOx 45.81 (0.74) 3.02 2.45 0.36 (0.17) PM2.5 1.76 (0.03) 0.10 0.07 0.00 (0.02) VOC 59.40 (0.98) 1.68 1.24 (3.42) (3.74) CO2 79,643.42 (1,403.50) (15.57) (982.75) (5,498.23) (6,427.33) Table 18. Tons of daily pollution emission reductions by scenario (compared with baseline absolute value). (7) (6) (5) (4) (3) (2) (1) - Smart Growth / Baseline Baseline / Roadway Smart Growth / Roadway Baseline / Smart Growth Smart Growth / Smart Growth Th ou sa nd s Figure 10. Daily reduction in tons of carbon dioxide emissions.

Modeling 61 transportation-investment scenarios, and, notably, the greatest gains appear to be when the smart-growth land-use scenario is coupled with the smart-growth transportation investments (transit and non-motorized). Assuming a value of $32 per ton of carbon—the highest value on the European exchange (PSRC 2009)—the smart-growth land-use and transportation- investment scenarios in combination would annually generate $53.5 million as compared with the baseline scenario, an increase of $7.7 million over implementing the smart-growth transpor- tation investments without the smart-growth land use. 8.4 Implications of the Modeling Results The results of the six model runs suggest that in addition to the social benefits that a smart- growth land-use configuration may have on passenger travel, there are also benefits directly to, and stemming from, goods movement. The largest benefits can be realized when a smart-growth land-use scenario is coupled with commensurate transit and non-motorized transportation investments. For trucking and shipping firms, the benefits include a reduction in overall travel distances and hours on the road, both of which result in lower costs. Secondary benefits, related to overall travel, include reductions in pollutant emissions, especially carbon dioxide, as well as the potential economic gains from a more efficient and productive goods-movement system. While great care was given to developing scenarios that could demonstrate the relationship between smart growth and goods movement, peculiarities about the Puget Sound region or the modeling tools used at PSRC may produce results that would be somewhat different under other circumstances. Because the State of Washington has had a Growth Management Act since 1990, requiring local comprehensive planning and transportation system concurrency, the baseline land-use and transportation networks may already, in some sense, represent a smart-growth regional pattern. To that end, modeling conducted with a sprawl-type land-use scenario for a baseline might show greater benefits than seen here. However, because the modeling described here did not show a general case, the benefits related to smart growth may not follow a linear function, and some levels of smart-growth land use, coupled with the appropriate land-use investment, may in fact show negative impacts. In other words, further research might show that there could be circumstances in which smart growth is too smart. Such research might include testing smart-growth scenarios in other jurisdictions using their regional modeling tools. There are also several reasons that the modeling results presented in this report may under- estimate the benefits of a smart-growth land-use configuration for goods movement now and in the future. As was discussed previously, current trucking models do a poor job of addressing truck trip generation, and four-step truck models do not include tours. As models more accu- rately depict the behavior of trucks and firms, they may estimate fewer trucks and shorter trips than is currently seen in the modeling results. Or, they may also be able to better distinguish between truck types, thus allowing for trade-offs to be made with smaller vehicles. However, in the absence of considerable data development and research to validate the improved models, it is unlikely that these models will be operational in the near term. Regardless, many important urban goods-movement issues (truck parking, delivery hours, etc.) are difficult to implement and perhaps are not often relevant for regional-scale modeling. In a longer time frame, the planning profession may begin to better connect the principles of smart growth to goods movement. Based on the focus-group results, it is clear that freight stake- holders would benefit from better relationships among themselves, particularly between public- sector planners and private shipping and logistics firms. As smart-growth developments mature to include further consideration of goods movement, perhaps better incorporating warehousing and distribution closer to urban centers or allowing for more flexible delivery modes and times, the benefits of smart growth for and from goods movement will likely increase.