Transportation and Land Use: Seattle vs. Portland
By Dick Nelson, Integrated Transport Research, Inc.
December 18, 2000
Portland and Seattle’s transportation and land use environments are often compared. Although Seattle’s Metro bus transit system has won national awards, Portland’s Tri-Met transit system is usually more favorably regarded because of its strong commitment to light rail, and its general commitment to melding transit and strict growth policies to conquer sprawl and the adverse impacts of the personal vehicle. In fact, Portland officialdom has used Seattle as the bad example it has chosen not to emulate. In response, Seattle has developed somewhat of an inferiority complex with respect to the transportation system of its southerly neighbor. But how well is Portland doing with respect to changing the public’s preference for the auto? Is Portland really winning the growth battle while Seattle is losing? This paper uses independent and reliable data sources to arrive at a tentative answer: Seattle and Portland are in the same (some would say leaking) boat.
Conclusions about Travel and Growth Patterns in Portland and Seattle
- Available data from reliable sources indicate that overall transit use and land use patterns in metropolitan Portland-Vancouver are not better than in the Seattle area.
- Even though Portland has light rail and Seattle doesn’t, overall transit use in Seattle and Portland is similar, about 56 trips per person per year in Seattle, and 58 in Portland.
- Automobile driving in daily miles per person is a little greater in Seattle, but is growing twice as fast in Portland and on a trend line to catch up. Government planning goals in Portland for reductions in driving have been scaled back.
- Traffic congestion in both metro areas is growing; freeway congestion is worse in Seattle, but arterial congestion is worse in Portland.
- Both Portland and Seattle have major air quality problems associated with pollutant emissions from motor vehicles.
- Urban centers designated for higher density and mixed-use development, (including transit-oriented development for bus and rail), show very little improvement in transit mode share across both Portland and Seattle.
- The federal government provided 80% of the construction cost of Portland’s light rail system, a share much higher than Seattle can expect to receive.
- Population per square mile in the urbanized area of Seattle used to be lower than in urban areas of Portland, but in recent years Seattle urban areas have become denser than in Portland.
- Population growth rates in the outlying counties surrounding both Portland and Seattle have been dramatically higher than growth in the urbanized center.
There are, of course, important related questions that cannot be answered without more focused research. We list some of these at the end of this paper.
Transit Ridership
Figure 1 is a chart from Tri-Met’s website. Tri-Met is the public transit agency for the three-county Portland metro area. It does not serve Clark County (Vancouver), Washington, which is the fastest-growing part of the greater Portland Metro area.
Figure 1. Portland Tri-Met’s Transit Ridership with the Opening of the Westside Segment of MAX
This chart is an example of a major problem one encounters in comparing transit system performance: what trips does one count? Tri-Met would like to take credit for a 55% one-year gain in transit use in the west side of it’s service area from the opening of the 18-mile MAX light rail line from Portland to Hillsboro. The reality is that Tri-Met in this instance is undoubtedly counting unlinked trips or boardings, and not actual (linked) trips, many of which involve a transfer from a bus to the train. As Ken Dueker, professor of urban studies at Portland State University has pointed out, the Tri-Met plan involves realignment of bus routes to feed the rail line, forcing bus riders to transfer at outlying MAX stations. So, although there may have been a growth in actual transit ridership, it was much less than indicated by the chart.
Another, and better, data source for ridership trends is the National Transit Database maintained by the Federal Transit Administration. Under federal law, all local transit agencies are required to annually submit uniform reports. The most recent data are for 1998. Table 1 compares ridership data per capita for King County Metro and Tri-Met. Again these are unlinked trips, the only ridership statistic available from the federal database. Also shown is the population and land area of the service area of each agency, and the total county area (one county in the case of Metro and three in the case of Tri-Met). The area used is an important number in determining transit ridership frequency. Under federal rules, transit agencies are free to define their service area as long as it complies with ADA rules.
Some agencies may show a relatively high frequency of ridership because they simply choose to serve a smaller and denser geographic area. You might call it cream skimming. Tri-Met, for example, serves just 14 % of the area and 72% of the population of the three counties (Clackamas, Multnomah, and Washington) centered on Portland. So its ridership appears relatively high when calculated as annual trips per capita. Metro, on the other hand, chooses to define as its service area the whole of King County. Its ridership appears somewhat more modest compared to Tri-Met. Yet when the total county population is used for both transit agencies, the trip frequency is not substantially different as indicated in Table 1.
Table 1. Comparison of Metro and Tri-Met Annual Trips per Capita
Transit Agency / 1998 Unlinked Trips / Service Area – sq. mi. / Service Area Population / Annual Trips / Capita - Service Area / County Area – sq. mi. / County Area Population / Annual Trips / Capita –County AreaMetro / 93,990,113 / 2,134 / 1,665,800 / 56.4 / 2,134 / 1,665,800 / 56.4
Tri-Met / 79,665,378 / 592 / 988,284 / 80.6 / 4,196 / 1,366,061 / 58.3
Roadway System
Given large geographical and topographical differences, meaningful comparisons of roadway systems in the Portland and Seattle metro areas can only be made with the coarsest of measures. Yet Portland and Seattle have some clear distinguishing characteristics, both dissimilarities and similarities. For example, Portland unlike Seattle has a freeway system that loops and provides easy access to its downtown. And with 12 bridges in a stretch of about 15 miles, the Willamette River does not present a bottleneck to east-west traffic flow that Lake Washington does with just two bridges in roughly the same distance.
Portland and Seattle have similar roadway capacity densities as summarized in Table 2, which shows selected federal-aid urbanized area data from the most recent USDOT/FHWA highway statistics report. Portland has fewer freeways but more arterial miles per capita.
Table 2. Roadway Densities in Portland and Seattle
Characteristic / Portland-Vancouver / SeattleTotal roadway miles per 1000 persons / 4.1 / 3.6
Miles of freeway per 1000 persons / 0.10 / 0.12
Lane miles of freeway per 1000 persons / 0.53 / 0.64
Miles of arterial per 1000 persons / 0.58 / 0.56
Auto Travel
The USDOT’s Federal Highway Administration requires that states annually report an estimate of the total vehicle-miles of travel on the roadway system in urbanized areas defined by the federal government for the distribution of highway funding. These numbers are obtained from loop detectors embedded in roadways that count vehicles. Table 3 provides daily vehicle miles of travel (DVMT) per capita for the Portland and Seattle areas for each of two years, 1993, and 1999, for which data was obtained in a consistent fashion. This recent data indicates that there is only a slight difference in the distance that residents of the two areas drive each day, while the rate of growth of personal travel is higher in Portland than Seattle.
Table 3. Comparison of Daily Vehicle Miles Traveled Per Capita in Portland and Seattle
Federal-Aid Urbanized Area / Total DVMT / DVMT/Capita1993 / 1999 / Change / 1993 / 1999 / Change
Portland-Vancouver / 26,049,000 / 30,985,000 / 18.9% / 20.3 / 23.2 / 14.3%
Seattle-Tacoma-Bremerton / 59,651,000 / 68,360,000 / 14.6% / 23.2 / 24.9 / 7.3%
Oregon Transportation Rule
In 1993, the Oregon Land Conservation and Development Commission, a state agency with the authority to implement the state’s growth management law, adopted what is known as the Oregon Transportation Planning Rule. The rule required Oregon’s four largest metropolitan areas to reduce vehicle miles traveled by 10 percent over the 20 years following adoption of a regional plan and 20 percent within 30 years. Reductions were also required in commercial parking space per capita. These reductions were expected to be produced through denser development around transit corridors and stations and through improved accessibility for pedestrian and bicycle traffic. The goal was subsequently scaled back to 15% in 30 years. Washington State has no similar rule.
Journey-to-Work Mode
As a part of each decennial census, the US Census Bureau asks how workers travel to work. Commute trip mode data for 1980 and 1990 are shown in Table 4. Both Portland and Seattle residents showed a significant decline in transit mode share over the ten-year period. The two metro areas exhibited very similar commuting mode patterns in 1990. Data from the year 2000 Census will be available in 2003 to allow a new comparison.
Table 4. Journey-to-Work Mode in 1980 and 1990 for Seattle and Portland
Mode / Seattle-Tacoma CMSA / Portland-Vancouver CMSA1980 / 1990 / 1980 / 1990
SOV / 73.5 / 73.8 / 73.0 / 73.5
Carpool / 9.7 / 12.3 / 10.3 / 11.9
Transit / 8.4 / 5.4 / 8.2 / 6.2
Walk / 4.2 / 3.3 / 4.7 / 3.5
Other / 2.0 / 1.4 / 1.7 / 1.5
Work at Home / 2.2 / 3.8 / 2.1 / 3.4
Congestion
The Texas Transportation Institute has for a number of years compared estimates of traffic congestion in urban areas. Congestion is measured in several different ways to provide and indication of trends for both freeways and principal arterials. It is well know that Seattle is has been among the top few urban areas in both total delay and cost of delay. What is not well know is that congestion on Portland’s roadway system has been steadily increasing, and that travelers on its arterial system experience more delay than do people traveling on Seattle’s arterials. These trends are shown in Table 5.
Table 5. Levels of Congestion on Portland and Seattle Freeways and Arterials
Urban Area / Congested % of Person-Miles TraveledFreeway / Principal Arterial
1982 / 1990 / 1996 / 1982 / 1990 / 1996
Portland-Vancouver / 30 / 35 / 55 / 60 / 60 / 70
Seattle-Everett / 40 / 70 / 80 / 50 / 55 / 60
Federal Share of Light Rail Construction Costs
Portland’s ability to afford the high cost of light rail transit was aided by generous federal funding. The Eastside line was funded in part by the by the transfer of federal roadway money. Both Eastside and Westside lines obtained much higher federal shares than is now possible as indicated in Table 6. The federal share of Seattle’s Link will probably decrease as the cost of the project increases beyond initial estimates
Table 6. Federal Funding for Portland’s MAX Light Rail in Comparison to Anticipated Federal Funding for Seattle’s Link Light Rail
System/segment / Cost / Federal shareEastside MAX / $321M (1986$) / 83%
Westside MAX / $964M (1996$) / 73%
Link / $3.6B (1995$) / 23%
Transit-Oriented Development
Portland has made a major effort to use its MAX system as a growth management tool, by seeking to increase residential density and mix of land uses near stations, and by employing various vehicle travel demand policies, including a state-imposed requirement that medium to large firms reduce the SOV commuting of their employees. The synergy between rail transit and land use, it is assumed, will reduce auto dependency and increase the share (mode split) of trips by transit and non-motorized modes. Seattle has embarked on a similar effort with policies to create higher densities and an appropriate mix of land uses around Link stations.
Since the eastside MAX line has been operating since 1986, there is some experience by which to gauge the effectiveness of Portland’s rail transit investment and supporting policies. Table 7 shows mode split and vehicle ownership data for four station areas on the MAX system, an area around a bus center, and two “town centers”. Town centers are designated areas for higher density and the encouragement of a greater mix of uses. Also shown is the mode spit for a Seattle “urban village”.
The Portland data was derived from a trip production-attraction model maintained by the City of Portland Department of Transportation. The numbers, which are for the year 1994, are percentages of all weekday trips that have an origin or destination within the defined center area, including trips that both originate and terminate within the area. Auto trips include both SOV trips and shared rides. Traffic that passes through the area is not counted. The Seattle data is from a 1992 travel survey of residents of the neighborhood center well served by bus transit carried out by faculty and students at the University of Washington. Again, the data is for weekday trips but does not include trips that originate outside the defined area.
Although the data sets for Portland and Seattle are not directly comparable, they individually show very little difference of transit mode share when centers, rail transit station areas and other centers served by bus transit, are compared with the region or city as a whole. The Portland data indicate a small tendency to increased transit mode share for centers that are closer to the central business district.
Table 7. Mode Split and Vehicle Ownership for Urban Centers, City and Region, Portland and Seattle - 1994
Type / Auto % / Transit % / School Bus % / Walk/Bike % / Vehicle Ownership% One or more
Portland Region / -- / 89 / 3 / 3 / 5
Portland Centers
60th / MAX station area / 86 / 6 / 1 / 7 / 93
82nd / MAX station area / 87 / 5 / 1 / 7 / 94
122nd / MAX station area / 90 / 3 / 3 / 4 / 96
148th / MAX station area / 84 / 4 / 7 / 5 / 99
Hillsdale / Bus transit center / 88 / 5 / 1 / 6 / 99
St. Johns / Town center / 90 / 4 / 1 / 5 / 93
West Portland / Town center / 90 / 4 / 1 / 5 / 98
Seattle City / 82 / 7 / 1 / 10
Queen Anne (Seattle) / Urban village / 72 / 7 / 1 / 20 / 1.74 avg.
Air Quality
Both Portland and Seattle have problems meeting federal air quality standards. These problems are in large measure associated with increasing volumes of vehicle traffic. Portland was recently found to be in noncompliance with clean air standards for ozone. Seattle has come close to exceeding ozone limits in the last few years, and long-range modeling suggests that carbon monoxide levels may eventually violate federal standards.
Urban Density Change
One way to gauge the effect of transportation and land use policies is to track trends in the spread of development and the density of urbanized areas. Every ten years, the Census Bureau defines boundaries of urbanized areas, measures their population, and calculates their population density. Table 8 compares the densities of the Portland and Seattle urbanized areas over the past 50 years. A new comparison will be available when the 2000 census numbers are released. These numbers clearly reflect the rush to the suburbs that began after WW II. They also indicate that, contrary to popular belief and regional planning agency reports, profligate land consumption in the Seattle metro area was not occurring in the 1980s. By 1990, infill was clearly beginning and population densities were increasing in both Portland and Seattle. The two patterns are strikingly similar.
Table 8. Population Density Trends in Portland and Seattle Urbanized Areas
Urbanized Area / 1950 / 1960 / 1970 / 1980 / 1990Portland-Vancouver / 5,057 / 3,626 (-28%) / 2,997 (-17%) / 2,869 (-4%) / 2,967 (+3%)
Seattle & Everett* / 4,517 / 3,387 (-25%) / 3,092 (-9%) / 2,940 (-5%) / 3,021 (+3%)
*Seattle and Everett urbanized areas were merged by the Census Bureau in 1970
Regional Growth
The following two maps show the percentage change in population growth in the last decade for the counties in the greater metro areas of Portland and Seattle. The Portland area extends from Vancouver, Washington (Clark County) south into the Willamette valley and comprises 11 counties. The Seattle area extends from Olympia in the south to the Canadian border, and covers 12 counties that fill the Puget Sound basin. Similar patterns of population dispersion are evident in both areas. Nine of the ten counties in the Portland area had population growth rates that exceeded the rate for the central county, Multnomah. One, Clark County, grew four times as fast and had the fastest growth of any county in Washington State in the 1990s. In the Seattle area, all other counties grew at faster rates than did the central county, King. Several of these were nearly triple that rate. These numbers indicate that population dispersion is a significant phenomenon in both extended metro areas, in spite of strong growth polices in both states.