Having reached the age of 80, one is tempted to look back at how places and people have changed. My appointment in geography at the University of Washington began in 1960, and I taught my last class in 2004. My main teaching and research interests were and are population and migration, urban and political geography, including social inequality.
Inspired by Michael Luis’ provocative Crosscut essay “Puget Sound Growth: where will they all live?", I offer this follow-up review of (1) population growth over 64 years, (2), the changing character of the population, 1970 and 2010, and (3) the planning context for growth. At the end, I will try to assess the roles of markets and of planning in this urban evolution — probably a good way to get everyone mad at me.
What would become the contemporary Seattle “urbanized area” in census geography was already a regional capital of almost 1.2 miilion in 1960, 864,000 in Seattle and 315,000 in the separate Tacoma urbanized area. But in the intervening 54 years, the population has far more than doubled, growing by 186 percent to almost 3.35 million — a story of growth more typical of the South and West than of the historic urban northeast. This growth was accommodated by an even greater increase in the physical spread of the urban area from 348 to 1,092 square miles, more than tripling.
Over these years the Seattle area has evolved, at times growing quickly, at times more slowly, as well as changing in the kinds of people, types of housing and patterns of settlement. It all occurred in a complex interplay of market forces (location of employment, housing preferences for locations and types of housing), and government — government at all levels, in its changing visions and regulations. I, a leftist (Scandinavian-type social democrat), see the world’s institutions as seriously flawed, the “market” subject to failure due to monopoly power, severe economic inequality and environmental harm. But, unfortunately, governments, in their attempts to address these problems are subject to comparable “planning” failures, in their zeal to control our behavior, as via “smart growth”.
The urban area population has grown by a remarkable 186 percent from 1960 to 2014, rising from 1.2 million to 3.3 million, and the amount of what is defined as the region's urbanized area has gone from 348 square miles to almost 1,100. The density has declined from 3,385 per square mile to 2,927 (the U.S. Census Bureau generally considers 1,000 per square mile or more as urban). The urban populations in 1950 were in separate urban areas, for Seattle and Tacoma, which did not merge until 1990. The Seattle area included not quite all the city, as well as most of Shoreline to the north, Highline and Renton to the south, and footholds across the lake in Kirkland, Bellevue and Mercer Island. The Tacoma area was mainly the city but included some of Lakewood and Parkland. The 1950s were “baby boomer” years of great growth in population and extent, Seattle expanding north to Edmonds and Lynwood, south to Des Moines and Kent, and east to include most of Bellevue. The Tacoma area grew mainly in Lakewood and University Place.
The 1960s were still years of high-family growth, economic expansion (e.g. Boeing) and high rural to urban migration, with the urban area growing even more than the population and leading to a large decline in density from 3,385 people per square mile to 2,906. By 2014 we still have not yet achieved the density of 1960 and before.
By 1970, the urban area extended north to encompass the city of Everett, and the corridor from Lynnwood to Bothell. Growth east now extended to Redmond, east to Lake Sammamish, southeast to Newport Hills and even Issaquah, and south to encompass a large area, including Kent Hill, Auburn, and parts of Federal Way. The Tacoma area extended east to include Puyallup, and southwest to add Fort Lewis and McChord. Although the areas now adjoined, they were not formally merged. The city of Seattle lost population, as it was relatively filled up, seemed out of date and less attractive to baby boom families desiring suburban houses and lifestyles.
Urban density areas of Puget Sound, 1950 to 2010
The 1970s decade was one of slower growth, from the Boeing recession as well as a drop in the birth rate, the “baby bust.” It was a bad era for core cities, and Seattle's population fell to 494000 in 1980, in part from white flight to escape school desegregation. Despite Seattle, overall density did rise in the region, and extension of the urban area was less dramatic than in the 1960s. To the north Marysville was added, to the southeast growth now spread toward Maple Valley, and for Tacoma, to Bonney Lake and into South Hill.
The rate of growth accelerated in the 1980s, a decade of both extensive suburbanization but the beginning of urban core recovery, especially in Seattle. To the north the urban area now added the Mill Creek-North Creek corridor, and reached Woodinville. To the east the urban area now included much of Sammamish, and Covington to Maple Valley. The Tacoma urban area moved farther into South Hill, east of Fort Lewis to Spanaway and across to Gig Harbor. The rate and extent of growth alarmed citizens, leading to pressures for stronger planning.
Rapid growth continued in the 1990s, again including a mix of both suburban extension and further urban core redevelopment, now advanced by planning programs. Nevertheless, overall urban density fell slightly and the areas of extension were quite spectacular. (although part of the apparent big spread was due to a very generous Census Bureau definition, sometimes dropping to include areas with only 500 people per square mile densities). In Snohomish County the urban area now added Arlington, far off Granite Falls (wrongly I believe), Lake Stevens to Snohomish, Maltby and on to Monroe. In King County, the urban area not only takes in Cottage Lake (OK) but even Duvall (well, maybe) , the rest of the Sammamish plateau, Mirrormont (dumb), Maple Valley, and Black Diamond. Extending from the Tacoma portion, the urban area encompasses South Prairie and on to Enumclaw (not convincingly), further south to Orting, and including much of the Gig Harbor area, even into Kitsap County.
For the period since 2000, the map showing both the amount and rate of growth is more helpful, and clearly reveals a quite mixed pattern of continuing far suburban growth, as on Union Hill, Snoqualmie, Bonney Lake and Orting, and Dupont, but also a lot of inner suburban intensification, and the well known densification of Seattle (south Lake Union, University district) and downtown Bellevue. Growth management and urban-center planning is making a mark, and leading to a rise in overall urban area density, but so far to a modest degree.
The year 1970 marked the end of a 20-year period of economic and population growth and expansion of the urban area. Those years, 19709 and 2010, are thus excellent years by which to trace the equally drastic change in the character of the population and the economy. Even in 1970, Seattle was relatively well off, not a downtrodden inner city.
Still in 1970 Seattle was a traditional central city in having a much higher non-white share, a far older population (median age of 33 versus an ultra low 25 elsewhere), a lower share of families and far more singles and roommates, but a higher share of female-headed families. The shares of single-family homes and of home ownership were lower than in the suburbs but were still a sizeable majority. As expected, transit use was much higher in the city, and the share of the college educated higher ( the UW effect). But incomes, home values and median rents were quite similar, or even higher in the city (and an example of the effect of Tacoma’s inclusion depressing the suburban numbers).
The 2010 city and suburban values show some similarities in pattern and some differences. Non-white and Hispanic shares are far higher than before, but now are lower in the city of Seattle (a reflection of gentrification). While in 2010 Seattle was still lower in shares of children, it was not much lower, and the median age actually slightly lower in Seattle, because of the large number of young adults in the city. Compared to 1970 the city is a little older, the suburbs are now a lot older, with a median age of 36.
Households have greatly changed from 1970 to 2010, but the relatively non-family character of the city remains. Seattle dropped to a family share of only 43 percent of households from 65 in 1970, while the suburbs dropped from an overwhelming 80 to a moderate 67 percent. The share of female-headed families, lowish in 1970, dropped even more in the city than in the suburbs (another gentrification indicator). Over the same period the shares of single person households, and especially of roommates and partners, has become astoundingly high, especially in Seattle. The popularity of tolerant Seattle for unmarried partners, the majority of whom are LGBTQ, is amazing, and is a somewhat distinctive gentrification measure. Suburban levels, while lower than in the city, are still high by national standards, and reflects especially the realization of an “urban growth center” character for Bellevue and Redmond.
The change in households is paralleled by changes in housing. The nature of housing in the city of Seattle has fundamentally and irrevocably changed, from a majority of family-owned single family homes to a majority of non-family apartment renters. But the shares of single family homes and of owners fell by a comparable degree in the suburbs as well as the city, and the share of single family homes in the suburbs is now lower than it was in the city in 1970.
The story of home values and rents in relation to incomes is fascinating. The values for Seattle and the suburbs were essentially the same in 1970 (but as noted above, King County and Snohomish County suburbs would have shown higher values than Seattle, about 15 percent more). Using the cost of living multipliers (2010 numbers are six times those for 1970), the most momentous changes are that median family incomes have risen only slightly, despite vast national growth in production (1.3 times in the city and 1.17 in the suburbs), rents have grown more quickly (1.6 times), but home values have risen extraordinarily (4 times in the city, 2.5 times in the suburbs). Dismal income values for families in 2010 simply reveal the well known decline of the middle classes, a national phenomenon. In the face of income weakness versus the high rates of increase in rents and, in particular, in home values, the issue of housing cost inflation cannot be avoided. (Tables that I developed on economic and housing changes are here and here.)
The shares of the population using public transit have risen, and remain far higher in the city than in the suburbs. The fundamental change in economic structure is seen in the shift from manufacturing industries to a much larger share of professional occupations, particularly in the city of Seattle, doubling from 28 to 56 perecent, with a concomitant increase in shares with a college education, tripling and standing significantly higher in Seattle than in the suburbs (56 percent to 40 percent).
To summarize, the greater Seattle-Tacoma metro area has more than doubled from 1.4 million to over 3 million over these 40 plus years. The urbanized area is “typical” in the sense that the majority (63 percent) are parts of families, live in suburbs (78), and in single-family homes (56 percent). But the ethos of our area seems to be set by the city of Seattle, espousing an urban future dominated by “smart growth” and interconnected by rail transit. But the city is simply different — unusually educated, professional and of higher class than its suburbs.
Washington state enabled comprehensive planning for cities and counties in 1958, and by 1970 Seattle and the three metropolitan counties — King, Pierce and Snohomish — had enacted comprehensive plans with modest attention to design and environmental concerns. Boundary review boards were able to ward off the most extreme and remote developments. But the sheer growth of population and jobs, and the anti-big-city sentiments of the 1970s and completion of the Interstate highway system created intense pressures for suburbanization of jobs as well as people. While there were large land-holdings (timber companies and railroads, for example), there were also large numbers of small holdings (one to 20 acres), many dating back to the 1920s. The rapid spread of housing into the “country” led to the perception and reality of excessive degrees of scattered housing and the loss of open space. This popular and political sense led to the creation of the Growth Strategies Commission in 1988 and enactment of Growth Management in 1990, the adoption of higher standards for subdivisions, development impact fees and the imposition of urban growth boundaries in 1992 (a fourth gentrification measure). The 1996 approval of Sound Transit ushered in the present era of strong planning, including concentration of jobs and people in designated centers.
The settlement impact of growth management is clearly seen in the map of population change, 2000-2013, in the prominence of higher growth in designated areas, and the relatively slow growth in the areas beyond the built-up urban area. But at the same time, growth also was high in several outer suburban areas, just within the urban growth boundaries, despite the recession.
I said I would trace the roles of markets and of planning in the evolution of the greater Seattle urban area. No one questions that housing costs are high (home prices and rents), and deemed unaffordable by ever larger numbers of households. Is it possible to assess the relative effect of market and of planning forces in bringing us to this state? An intriguing sub-question is of the special nature of Seattle: Why is it so different? Can it be changed? Well, I taught statistics and am well-trained in urban and regional economics, but I do not feel confident in definitively answering the basic question.
Markets created the demand for both houses and apartments, depending on the kinds of households, income and the kinds and locations of jobs — which, in turn, are mainly market-driven. Planning — through urban-growth boundaries, impact fees and selective up-zoning — has revealed predictable preferences for higher density and shares of apartments. These, but especially urban growth boundaries, clearly impact housing costs, through raising the cost of land (supply and demand). But this effect, overall, is almost entirely a suburban phenomenon, since there was essentially no remaining land in Seattle for urban expansion. Plausibly, growth boundaries raise the cost of land 50 percent or more.
A significant market effect, especially in the city of Seattle, has been the sheer numbers of childless singles and couples, creating a high demand for central living and for apartments, to the joy of planners.
Growth has helped change Seattle's skyline Credit: JoeinSouthernCA/Flickr
But the most important market effect on the nature of settlement and kinds of housing offered is the rising inequality of income and the concentration of wealth in the perhaps top 5 percent, which has severely impacted the capacity of the remnant middle classes to compete in the housing market. This is not the fault of planning, but of the economy.
Another market effect of the growing power of the richest 5 percent (and of higher shares of young singles and partners) is altering the patterns of settlement, as they have the power to claim or reclaim areas deemed desirable. This basically answers the second question: Seattle has such high appeal that values and prices of housing are inflated to very high levels, and far more a market effect than a consequence of planning. Of course, Seattle could decline in a great recession or if Amazon and the like were to leave the city, but I would not bet on that.
The housing cost inflation effect of growth management is significant and could lead to some shifts of employment as well as population out to less restrictive cities and counties, as the power of planners is probably too high to consider a needed loosening of the urban growth boundaries. But for far into the future, income inequality is the stronger market force in compelling a shift from single-family homes and ownership to apartments and renting. So ironically, smart growth results are as much or more a result of market forces as of planning.
But the market, somewhat aided by planning, does afford some creative alternatives to apartments, and especially apodments. The growing prevalence of “accessory units” is not popular, but probably realistic in the face of the popularity of living in Seattle. Even more prevalent, and acceptable are the twin phenomena of adult children returning home, or of elderly folks moving in with their kids, or the housing of more than one household in what only technically were viewed as single-family. This is what might be called the “Los Angeles” solution, as it pioneered there long since. This “doubling up” in traditional homes at least affords the highly desired private space for families, and allows the population to increase, city and suburban, without depending so much on apartment construction. Another strategy is simply to return to smaller lots and even attached row houses, again affording private space.
The very poor and the homeless are outside the market and only marginally admitted by planning, instead relying on direct political intervention that is outside the normal planning process and outside the housing market.
Markets and planning are imperfect. The evolution of the Seattle urban region is a product of both. Before 1990 planning did influence the broad patterns of job location, and housing of different kinds, but the market mainly dictated the kinds of housing and popularity of locations. Since 1995 or 2000, planning increasingly has controlled location more strongly and the types significantly, for good or ill.