Beloved and Abandoned: A Platting Named Portland

For American planners, Portland, OR is held up as a shining example of urban planning, and credit is given to its compact grid. But is Portland's grid worthy of adulation? Perhaps not, say Fanis Grammenos and Douglas Pollard of Urban Pattern Associates.

In the 4000-year history of the grid, American incarnations are relatively new, appearing first about 300 years ago, frequently as a simple, orthogonal and often square (such as Portland's) 'Hippodamian' grid, named after the planner of Miletus around 473 BC (Fig 1).

A session in the recent 2009 New Partners for Smart Growth conference focused on 'The Beautiful American Grid - the Embodiment of Smart Growth,' which lamented the fact that the grid 'gets no respect'.

Figure 1. From Miletus (473 BC) to Barcelona (1859) and to Portland (1867), the grid has resurfaced in many alternative sizes of square blocks and in variations of elongated rectangles (NY 1811).

This alleged lack of respect seems at odds with most planning literature, which extols its virtues and mirrors prevalent New Urbanist practice. This disparity between theory and practice simultaneously confuses the practitioner and frustrates the theoretician. It deserves detailed attention if only to clarify this schism and enable site plan designers to know when and why they could apply 'the Grid'. Clarity about its attributes may also open the way for its regeneration.

Recognition and Respect
Current planning literature brims with references to "the grid" in juxtaposition with curvilinear and dendrite conventional suburban layouts. The "grid" as a network concept has been widely accepted and is now regarded as a superior geometry for laying out greenfield and infill sites.

Figure 2. Portland's (Hippodamian) Grid overlaid on a Virtual Earth bird's eye view of Pearl District. The centre lines of streets intersect at 260 foot intervals.

For example, in 1992 we read that "Streets ought to be laid out largely in straight segments, as they were until the 1940s. After all, the vast majority of our successful towns and cities, from Cambridge to Portland, were laid out this way." (Duany). The grid gets credit for city success, at least by inference, but is this credit warranted?

Portland's network offers an instructive example for discussing grids because of the grid's nature (an unadulterated Hippodamian grid and the densest of all American city grids (Fig 2, MS Earth), its size and the City's planning celebrity status. We read again that "Portland owes much of its success to its tiny blocks that create an incredibly porous network of streets, each of which can be quite small as a result" ( Jeff Speck, 2005). In this praise, it is not simply the grid in general, but the small blocks in particular that impart success.

In articles, project brochures and city planning reports "the grid" stands alone; the other contestant, mid-to-late 1900s suburban networks, has been wholly discredited in mainstream planning. One can hardly pay more respect.

Affiliation and Affection
Portland's street grid pattern has attracted attention indirectly and directly. Indirectly, because the City of Portland has taken many first-ever, brave and decisive measures to manage growth, and cities and planners hold it up as a model of civic vision. Inevitably, attributes of the city -- such as its grid -- are seen by affiliation as paradigmatic.

Personal testimonies of visiting planners who express adulation for Portland add a second indirect layer of attention. Constantly on the outlook for an ideal urban pattern, planners list Portland as a favorite and some boast "I love that city!" with emotion. Recently, a local movement to rename the city in order to project these strong emotions was set in motion. But strong feelings such as these may be entangled between actual attributes and personal associations, hard to unravel for practical purposes, as other cities also share such emotional investment, at times.

Urbanists and romanticists have expressed equally strong sentiments about Paris, London, Barcelona, Curitiba, Amsterdam and Venice. Of the six, only Barcelona adopted the Hippodamian grid in 1859 for its vast expansion, and Venice, without a classic grid, is the preeminent pedestrian haven, yet neither city matches the urbanist's praise for Portland. Whatever the mix of reasons, Portland dominates the American planners' imagination feelings and talk. Disentangling this intangible realm can be an elusive goal; grounds and figures on the other hand may produce tangible results.

Grounds and Figures
Pragmatic reasons may play a part in this adoration. The extreme simplicity of the plan, for example -- a uniform, perfectly orthogonal, expandable checkerboard -- could be one. As a drawing, the plan has a feel of flawlessness, the appearance of perfection, particularly in contrast with labyrinthine medieval town plans or recent bewildering suburbs (Figure 3). When this perfection is combined with a pleasant experience on the ground an indissoluble match is made.

Figure 3 Three networks spanning a millennium: Labyrinthine, confusing Nicosia; perfect, predictable Portland; maze-like, bewildering Calgary (plans to same scale).

The degree of connectivity of the street network could count as another practical reason. 'Network', by definition, is a set of linked components, whether a spider-net, a fishnet, or the Internet - all networks connect. What distinguishes them is the manner, geometry and frequency of connection: leaf, tree, blood vessels, telephone and web networks are dendrite, hierarchical (fractal) but fishnets are not. Portland's is a dense fishnet with nodes at every 200 feet, which produce 360 intersections per square mile -- the highest ratio in America, and 3 to 5 times higher than current developments. For example, older and newer areas in Toronto, typical of most cities, range from 72 to 119 intersections per square mile in suburbs and 163 to 190 in older areas with a grid. As connectivity rose in importance as a planning principle, Portland's grid emerges as a supreme example.

Coupled with connectivity, its rectilinear geometry is indisputably more advantageous for navigation on foot, car or bike than any alternatives. Visitors often feel lost and disoriented in medieval towns and in contemporary suburbs and this feeling leads to anxiety and even fear and a sense that all is not well.

What explains why the simplest, purest, most interconnected and easily navigated rectilinear grid, in spite of all the praise, has, evidently, not been applied in any contemporary urbanist plan, whether infill or greenfield? What caused the disaffection?

The Disaffection: Speculation
One clue comes out of a believable legend about Portland's grid. Unlike other American cities that were laid out by erudite generals or governors, such as Oglethorpe (Savannah, 1735) or William Penn (Philadelphia, 1701), Portland's plan was apparently conceived by scrupulous speculators who reasoned that more corner lots would yield higher profit on the land investment, hence the maximum number of intersections. Interestingly, the 1812 Commissioners Plan for New York was also denigrated as a 'speculator grid'. The 'speculator' label would usually damage the prospects of any plan; speculation is perceived as shortsighted, greedy, and at times suspect activity -- as opposed to "planning" which is a long term, public-good, goal-centered activity.

Interestingly, a more contemporary "speculative" calculation may be the equally pragmatic reason for its abandonment. The Portland grid uses 42% of land in right of ways for streets and has the highest length of road infrastructure of any alternatives. Simply put, nearly half of the land is used up in accessing the other half. A recent comparison of an existing 338 hectare subdivision's curvilinear pattern to an overlaid TND plan showed that the land for roads was respectively 88 and 122 hectares or 40% higher for TND with a corresponding increase in infrastructure costs (IBI) (Figure 4). No developer or municipality would savor this arithmetic.

In business districts, small blocks may force buildings to gain height and thus increase the per block net density, a financial advantage, but the gross density of such district would be comparatively lower than that of another with larger blocks and similarly tall buildings. On balance, more buildable land means more opportunities to build, tall or otherwise, and therefore more rentable space, revenue and activity.

Evidently, Portland's founders either understood little about infrastructure costs or judged them irrelevant; a judgment that no planner, developer or municipality today would take at face value. When economic efficiency matters, Portland's grid fails the grade.

Figure 4. Comparative Building Block sizes of Portland, Suburb and Suburb TND (partial plans). (Note the total eclipse of 4-way intersections in both newer plans).

Reasons that relate to urban design aesthetics can also be seen as contributing to the disaffection with Portland's platting. Starting with Camilo Sitte in 1892, who said categorically: "Artistically speaking, not one of them [grid patterns] is of any interest, for in their veins pulses not a single drop of artistic blood." The string of unfavourable comments continues to 1994 with "Upon reflection, we realized that the developers [who hired us] had a valid concern, one related to the shopping-center developers' understanding that human beings do not like endless vistas." (Duany).This insight into people's behaviour was confirmed by academic research recently (Ewing). Add to this backdrop the common, if superficial, perception of cookie-cutter planning and endless monotony, and distaste for the Portland grid emerges, particularly in eclectic urban designers.

Since Ian McHarg's 1969 classic book, Design with Nature, planners have been keenly conscious of the potential negative impact of land development on natural systems. Soon after, pioneering projects, such as Village Homes (1975), responded to this concern. Recently we heard: "The New Urbanism does not do grids that quash nature" (Duany 2001) followed by a movement for Low Imprint New Urbanism in 2007 (LINU). Permeability and rain water management have emerged as key indicators of a plan's fitness. On these measures, the Portland grid occupies the negative end of the spectrum of impermeability with the most road surface. With environmental concerns and regulations rising to the top of the planning agenda, any low performance plan would be disfavoured.

Compact, dense development, such as happens in downtowns, lowers the pressure for expansion and its incursion on natural environments. However, though a city's bioregion may be better off, the dense downtowns still exports large amounts of storm water and, with it, pollution. No part of the city need be absolved of the imperative to curb outflow; greening unnecessary asphalt is a viable first step. In that vein, Portland has retrofitted some streets.

Safety and mobility
Practical considerations about traffic flow and safety may also undermine its presence in contemporary plans. The term 'gridlock' fixed in the planner's vocabulary the sudden realization that the grid and car traffic may, at times, be wholly incompatible and that the conflict increases with the grid's density, as the space for stacking diminishes. The alternative to the grid, 3-way intersections, has been established as the safest and as enabling good flow. (Lovegrove, IBI). When streets in a grid become alternating one-ways, as in most downtowns, they create virtual 3-way intersections throughout an entire district, and achieve both safety and flow. Virtual 3-ways result also from traffic circles, as in Seattle and Vancouver, and from roundabouts, now gaining acceptance in America.

Figure 5. Diagram in ITE's 1999, advisory document on TND street network planning.

The ordinary impression on the ground that the Portland grid 'works' in contemporary traffic conditions is casually taken as a sign of suitability. This view obscures an entire century of engineered physical, mechanical and management adaptations which are overlaid on the 1866 platting. Remove these (in a thought experiment) and imagine the outcome. Clearly, an ill-suited geometry is made to work with interventions such as dividing lines, medians, traffic signs, traffic lights, directional signs, bollards, street widening, one-ways, traffic circles or roundabouts and many others.

Abandoning the Grid
The current map of Portland shows the transformations the city's grid has gone through since the 1866 platting, a century before environmental and traffic issues drew the spotlight.

In the car-less world of 1891, a variation called ‘Ladd's neighborhood' was built, ignoring the surrounding perfect grid and follows a Beaux-Arts, L'Enfant-inspired plan with diagonal streets, (Figure 5) disrupting it.

Figure 6. Three layouts showing the departure from the idea of the 'grid' (all plans to same scale)

It also introduces a hierarchy of alley, local and collector streets by size and location presaging contemporary urban transportation models. In a sea of formless, perfect uniformity, it brings an organizing module (about 160 acres) that anticipates Perry's Neighbourhood Unit (1923), which also assigns a hierarchy to its streets, and, likewise, protects it from through traffic.

Transformations also happened within and beyond the 1866 city outline over time: blocks doubled or tripled in length, some streets became discontinuous and, later, curvilinear streets appeared. More recently, some of the city streets were closed to cars, effectively doubling the block size and introducing a pedestrian space in the middle; an adaptation that produces a high quality public realm which is in short supply in an extensively asphalted grid. All these transformations occurring next to an "ideal" grid leave a trail of desertion which is hard to reconcile with the affection found in literature.

For reasons of land efficiency, infrastructure cost, municipal expenses, rainwater management, traffic safety and flow, and the demand for increased pedestrian share of public space, the praised, pure Portland platting will likely not find new followers.

Portland will remain a adored and beloved by urbanists, but her Hippodamian grid layout seems destined for the archives, abandoned as a good idea of a byegone era. This transcendence leaves urbanists, who seek to regenerate a contemporary urban pattern that is as pure, complete and systematic, looking for alternatives: ones which excite the same first blush of adoration and delight and lead to a deep abiding love, but also hold up to intense scrutiny of their economic, social and environmental performance.

Fanis Grammenos is a principal of Urban Pattern Associates and was a Senior Researcher at Canada Mortgage and Housing Corporation for over 20 years. He focused on housing affordability, building adaptability, municipal regulations, sustainable development and, recently, on street network patterns. Prior to that he was a housing developer. He holds a degree in Architecture from the U of Waterloo.

Douglas Pollard is a Senior Researcher at the Canada Mortgage and Housing Corporation.




I am a Planner from New York City. My first Porland experience was in 1968 for the 100th meeting of the AIA. I was there to hear Whitney M. Young. You know what happend in NYC since then.

I am convinced that confidence in the city's capacity for innovation is embedded in the constraint of the grid. Speculative in origin or not, too costly to repeat or not, it serves the ability of the small investor or large to believe in renewal and to reclaim a lost block or building.

Take another look at your history books, places without grids fail to produce dense centers and block to block diversity in land use or culture. They are hard to find because they are ghost-towns.

Start thinking a little past the classic amortization period and see a chain of investors, not just the spread in the present.

Success by Platting

Planners would unreservedly emulate proven success, particularly if it can be achieved with means as simple as the Portland grid. However, “success”, a fuzzy term, is evidently difficult to achieve. In a study that looked at mid-sized cities downtowns in the US and Canada, the vast majority of which are laid out on a grid, we read: “The downtowns of only 19 out of the 202 listed in the questionnaire were placed in that group [the successful few]”. (Piere Filion). Curiously, one of the 19 is not a grid. Also, some large cities, such as Detroit and St Louis, for example, are proof of a misplaced emphasis on the grid as a source of success.
By contrast, many European, African, Middle Eastern and Asian cities have central districts that originate in the middle ages or earlier and display a geometry as far removed from the pure Portland grid as one can get. Yet they have been very successful. It seems that many other forces than simply the layout geometry support or undermine success.

Forces that determine success

What are the major forces aside from layout that you believe determine success?

I can see the pros to the grid layout. However, I also appreciate places in cities where the grid pattern is broken up with wider street spaces and v shaped common green spaces/mini-parks in the middle, allowing for spaces that feel open and unconstrained by the boxy feeling of a grid.

City Fortunes

Cities have seen their fortunes rise and dwindle and so did peoples' lives in them. Same layout, different state of being along their historical trajectory. "Success" has to be broken down to specific criteria and metrics (leave the glamour aside for a while). A complicated task, but some are trying. Health, for example, could be one. The article I mentioned above (Pierre Filion) lists many more.

the grid

What explains why the simplest, purest, most interconnected and easily navigated rectilinear grid, in spite of all the praise, has, evidently, not been applied in any contemporary urbanist plan, whether infill or greenfield?

except for the infill neighborhood in figure 2 above. thats the new hoyt street yards in portland's pearl district and it didnt exist 10 years ago. it just continued the surrounding portland street grid through the former rail yards.

Research re Humans Disliking Endless Vistas


Can the author explain the subject of research demonstrating that humans do not like endless vistas? What is this in respect to?

I quite enjoy endless vistas of natural areas. Does this pertain to lack of a feeling of enclosure in built environments?

Please expand on the concept and the research.

Thank you.


It does have to do with built environments.
The topic of enclosure and its analysis becomes quickly abstruse and the realm of only few specialists. Most cities were laid out without any conception of whether such an issue existed, particularly the ones following the grid. The only true enclosure early city dwellers understood well was the life-and-death matter of fortification – not a subject for aesthetes or philosophers.
My first encounter with the innate desirability of enclosure was in a 50s text praising the cul-de-sac, the ultimate street enclosure; the statement was conjectural and gave no research references. The second, was in the reference mentioned in the article, where the knowledge is attributed to developers of shopping centres, also conjectural. Finally a recent academic article dealt with this and other topics of “likes” and “dislikes” in a more rigorous way.( Identifying and Measuring Urban Design Qualities Related to Walkability by Reid Ewing, Susan Handy, Ross C. Brownson)
While enclosure could be a design consideration, it most likely would not top the list of concerns about making neighbourhoods more enjoyable.

Late Adherence

It is instructive to know that there is at least one recent exception to the rule of not following the Portland grid. It would be even more instructive to understand whether this late adherence was driven by planning, transportation, speculation, or environmental priorities and what these were. Considering that the intentions and ideas of many town founders are generally lost the further one moves into the past, it would be doubly important to record recent, traceable priorities so that we can better understand what we see on the ground. And since the Pearl and surrounding districts display several street closures and block consolidations (recent?), it would be also informative to contrast the logic that enabled these.

Consider the World Trade Center

Recent planning at the World Trade Center, while not dealing specifically with the Portland street grid, may help address some of this interesting question.

When discussions began about rebuilding the World Trade Center, consensus quickly emerged that the "super block" created by eliminating streets was an undesirable urban form. Planners and citizens alike generally agreed that some of the (somewhat irregular) street grid should be extended back through the site.

The reintroduction of two main streets, Fulton and Greenwich, was quickly adopted as the universal consensus. Other, smaller streets (Cortlandt and Dey) became points of contention.

The site owner and developer did not want to include the additional streets, based almost entirely on their development needs. In order to provide room for the large building cores for tall buildings and still have marketable space on the lower floors, the developer would not agree to the smaller lot sizes.

Meanwhile, the City's planners, with backing from a generally pro-developer Mayor's Office, were pushing to put the full grid streets through the (non-Memorial portions of the) site. They provided a whole range of massing and street life arguments. The general theory was that the smaller blocks were more pedestrian-friendly.

It appears that, at least for now, the developer has prevailed. In this case, that is probably because the site owner is an insulated, separate branch of government. Had the Mayor's Office had authority, the small-block grid would have been restored to the World Trade Center site.

Constructing a straw man

I flew out of Portland this morning. I think we need to clear up some misunderstanding or misrepresentation regarding the Portland grid in this article. The facts on the ground are certainly not "an unadulterated Hippodamian grid", and the figure pictured above appears to be nothing other than fiction.

Portland has more than one grid, even in its oldest and most central areas. Two grid patterns that are oriented differently collide along Burnside, producing a series of oddly shaped triangular lots and buildings.

I find it difficult to believe that anybody who has actually visited Portland could honestly even suggest there is a "perception of cookie-cutter planning and endless monotony". This would require ignoring topography, public art installations, public spaces that frequently interrupt the grid, and the already-mentioned lack of a truly continuous grid.

The concern about efficiency is intruiging, but an incomplete analysis at best. While it is true that smaller blocks do consume more land for streets, it is also true that the amount of street parking also increases. Without providing a more sophisiticated look at the tradeoffs of on-street and off-street parking - which should also consider the quality of the urban design - this analysis presents a partial picture that only presents the negatives.

Less parochially, we should also recognize that in places like Portland, the streets truly are public space. To make this point bluntly, minimizing the streets for more private development would be nothing other than replacing the commons with more shopping malls.

The "gridlock" concern is misguided, and ignores much of the rationale frequently cited to support street grid patterns today. This point should have been dealt with more fairly. The authors must be aware of the line of argument that street grids provide redundancy that limits the congestion that occurs when everything is collected onto arterials - especially if there is any sort of incident that affects one of the arterials.

We must also reject the argument that the Portland grid is inadequate because of the use of traffic engineering methods to optimize its operation. If we were to accept this standard, there would really be no functional dense urban traffic patterns at all. Of course, it is largely the traffic engineering magic of paired one-way streets creating signal progression that makes grid patterns work so well (especially with the redudancy already mentioned). In fact, the suggestion that the use of traffic engineering is somehow illegitimate serves only to obscure just how flexibly the Portland street pattern has adapted to changing modes of travel and levels of demand.

Expanding on this point, the comparison of successful non-grid city centers to Portland misses much of the point of how well Portland functions. Even while the Pearl District seems to have become an expensive yuppy shopping area, Portland still works for a wide range of economic purposes. It is not a downtown-cum-boutique for tourists. Moreover, the efficiency of making deliveries and meeting all the other modern urban needs helps keep down the cost of living. Downtown Portland does much more than Venice (and does it well), so let's compare apples to apples!

Finally, this article seems to completely ignore the real reason grids are so rarely introduced now: cities no longer develop large-scale plans for their street network expansions at levels smaller than arterials. Portland may have planned a grid to serve the interests of developers, but it was coordinated and planned in advance. Today, developers have control over platting at the level of individual projects, which yields to local site conditions and bends to maximize the value of a handful of lots. The loss of the grid is partially due to the loss of master planning. As a small town planner in Oregon told me this week "People don't want the government to have that much control." You would not see a grid today, because it simply wouldn't be the outcome of the individual interests of the separate developers creating plats on adjacent properties at different times.

The Portland street pattern is not perfect. Unfortunately, we are given some strange fiction about a monotonous stretch of streets that doesn't even exist, instead of a thoughtful critique of any of its actual operations.

Facts, not affectations

We agree with author’s conclusion that “the Portland street pattern is not perfect” but find no evidence in the text to justify this statement. Also there is little that shows why the Portland grid is good, apart from generalities. By contrast, our article showed how the grid fails based on facts or on current accepted opinion. We will follow the author’s thread to once again reinforce the facts.

Variety of Grids
All cities are patchworks of layouts; that’s evolution and adaptation; not the point. The article refers to the famous and admired Portland grid shown in the 1866 platting that can be seen in a City website:
This map shows a uniform grid east of the river extending in both directions and the exact same grid on the west side with one change in direction. The question of why the “good” grid has not been replicated remains a challenge.

We fully subscribe to the view that a regimented layout does not necessarily yield a boring city, which is implied by our “if simplistic” clause. We also believe that the geometry on paper rarely says much about the experience on the ground. It appears that respected contemporary planners view the Portland-type grid as monotonous and unappealing. We are not defending such views; we simply state that they exist and that they could be part of the nexus that explains the disregard of it and the embracing of the “village” picturesque. The sources are noted.

Efficiency of Parking
In the historical vein of the article, Portland’s founders knew nothing about street parking; a future priority, among others, that they missed. We now find their plan convenient for parking; an afterthought, an adaptation. This presumed contemporary advantage has not lead to its replication in New Urbanist plans, an intriguing absence. One explanation might be that parking is in high demand only downtown, not in the suburbs or exurbs where most New Urbanist developments are found.

But this logic runs into trouble. We imagine downtowns as primarily pedestrian places hence the proliferation of pedestrianized city cores and streets, gallerias and lanes, downtown shopping complexes with multilevel internal streets and the current trend of charging drivers to enter the downtown. The urbanist philosophy tends to say: get on your feet, take transit, rather than parking is available for you if you prefer driving everywhere.

The above points would be supplementary if the parking arithmetic worked. But it doesn’t. Joining two blocks, as was done in Portland’s NW district, yields at least 10 new parking spaces and a 12,000 sq foot lot (for apartments or offices), in the place of 8, a gain of 2, assuming one-way traffic and full use. The urbanist choice is obvious: More space for people and for cars. More people downtown, fewer drivers.

That’s not all the arithmetic, either. Assuming that the in-between street remains a ROW, an open space, we can park 14 cars total on the first 20 feet of it on either side and regain the four previously lost to turning motions. Increased parking, more open space less asphalt; a no brainer.

All this accounting pales when the parking numbers are listed: 4,000 city owned spaces and as many or more in private garages. Urbane cities, like Portland put their parking in mixed use buildings under or above ground.

Clearly, the parking and development arithmetic does not favour the Portland grid: Not downtown; the more valuable the land the less reason to use it for parking. Neither does it favour it in the periphery; less need for roads and parking and greater need for lots and common spaces. Street parking is the Small American Town solution not metro-USA.

Commons trade-off
The real trade-off is between types of commons: Streets vs squares, parks, churches, institutions, stadiums, galleries, museums, movie theatres, the opera and pubs. All of these “commons” need land downtown, scarce land and some of them cannot fit on a Portland block. For example the PGE park and the Rose Garden Arena each require 6 Portland blocks. To label downtown streets with more than 10,000 VPD “public domain” would be to ignore noise, exhaust, risk and the health and frustration of pedestrians. It also ignores the new streets where the “public” dominates such as the Stroget, Venice or Vauban type; only people and no cars in sight. (

Gridlock Logic
Had the traffic advantage rationale been persuasive, we would see more Portland grids in new urban plans; but we don’t. This may be because the advice is too general or contradictory. Grids can be conceived and constructed at many scales: the block, the collector, the arterial and the highway scale. At each scale there are observable thresholds beyond which the system fails. Without specifics, the principle is freely interpreted and of little practical value; any grid is a good grid, therefore, why Portland’s. But the gridlock arithmetic suggests that while Portland can stack 8 cars at intersections, Sacramento, Barcelona and NY (in cases) can each stack 18, 20 and 30 respectively. On a gridlock arithmetic Portland loses out.
The contradictions in the guidance can also discourage planners. In the ITE guide we read: “While TND networks do not follow the same rigid functional classification of conventional neighbourhoods with local, collector, arterial and other streets, TND streets are hierarchical to facilitate necessary movements”!. And if local streets are made to T intersect, as recommended, traffic dispersal cannot be achieved.

Evolutionary Perspective
Engineering interventions are welcome and inevitable but make a point from an evolutionary perspective. Humans and most creatures adapt; humans through ingenuity and creatures through genes. Adapting an inherited street pattern to new conditions is a survival tactic, but it also proves that the pattern needs adaptations to function well. But humans have the gift of foresight; they can think from first principles, understand why a system malfunctions and redesign it so that it would not; designing the future not just fixing the past. The key point about adaptations then is a “what if” question: IF Portland’s founders knew the car was coming, would they have laid it out exactly the same way? And for contemporary planners: knowing what a car is and does, would they use the Portland grid to solve for the variable “traffic”? Few have chosen to do so not even in Portland.

The apples of comparison
But what are the apples of comparison? Economic power, jobs, equity, affordable living, street efficiency, standard of living, shopping etc? Venice used to be a world power, now a world boutique (but still a pedestrian haven) and Detroit used to be the industrial powerhouse of US, now a city in distress; none of that tells us anything about how to lay out a neighbourhood. And do the apples have something to do with the grid?

Master planning cul-de-sac
Imagining Portland’s founders as master planners is a stretch. Contemporary master planners in Cities, admittedly, create policies but have little control over the physical form. Yet, there have been many occasions where developers (contrary to their vilified image) have given practicing planners an opportunity to do master plans on a large scale (defunct airports for example) upwards of 3,000 acres. None of these applied the Portland grid. Is the absence of evidence, evidence of the absence of the worth in doing it?

The imperfections
If not its land inefficiency, its traffic unsuitability, its negative environmental impact, we wonder what the author had in mind with the phrase “the Portland street pattern is not perfect”.

When talking about streets, transportation does matter

There seems to be some disparity in what is meant by "street grid" in this discussion. The authors seem to have somehow tilted the meaning away from the network of public transportation right-of-way to mean something much more narrowly conceived as the automobile lanes alone. If that is, in fact, the case, some of the problems with their arguments may fall off. On the other hand, it would seem to be picking and choosing the definitions to suite their purposes, and it would fail to recognize the multiple purposes of true public streets, and their need for flexibility over time.

Streets are public spaces, but they are not parks. And once you transform a portion of them into a formal park it no longer has the flexibility to serve future changes in transportation needs. If you surrender streets for building sites, as these authors seem to encourage, all flexibility is gone!

But let's start with the main point first:
What evidence is there, of any form, to actually support the theory of some "traffic unsuitability" of Portland's downtown grid. Are there examples of intersections operating at a poor level of service? Can you point to any queuing analysis that shows the block length cannot be reconciled with the signal timing? As already stated, the grid seems to have shown a remarkable flexibility to adapt to new modes of transportation and a surprising level of growth.

Beyond that, let's try to stick to some basics about transportation in cities. I appreciate the example of storage capacity along the block, and it may actually provide some real reason to think the block size in the Portland grid may be too small. However, without actually examining the volumes and signal progression, this is merely conjecture. And in this case, it seems to be conjecture that does not match the experience on the street.

The authors cited the notion that there is value in corner lots, and then they ignored it. They want to say small blocks are not meaningful, but then they fail to address the very reason they say they were created. Is this value unimportant?

Perhaps they are unaware of the way this same principle was even used previously to argue against paired one-way streets, finding that they reduced value by reducing the visibility of the corners not seen by approaching drivers. Of course, this would apply less in a transit and pedestrian-rich place like downtown Portland, where potential customers are not so auto-dependent.

The authors apparently choose to ignore the theory proposed by Jane Jacobs that small blocks provide more points for pedestrians to intersect, increasing sociability. They surely are aware of this. Unfortunately, they seem to overlook or too easily dismiss these ideas about how people use and enjoy streets.

Even with very low auto usage, the idea that transit can replace all parking is simplistic rhetoric, and a mistake often made by planners with little experience in the field of transportation. Not all vehicular trips are made for personal trips, and parking is necessary in business districts for the many deliveries, pickups, and service trips made throughout the day. Using off-street parking for these purposes adds to the cost of doing business, if you consider how the additional time for each stop can add up over the course of a business day.

It is simply an implausible argument that we invest poorly when we have a full network of public streets with on-street parking, as this space efficiently combines an effective sky exposure with less expense for structured parking. Interestingly, the authors seem to ignore the coherent arguments that have been made so frequently about the value of parked cars as separation to shield pedestrians from the nuisances of moving traffic on the street, and about their role in calming traffic. This is precisely why many Oregon cities are now encouraging on-street parking after decades of growth without it.

Of course, if you take the narrow perspective of developers, these public benefits may appear to be merely extra cost or "inefficient", and this is one important reason why grids have not been extensively replicated today.

Before setting aside on-street parking, though, I must recognize that it is problematic for bicycle transportation. Dooring is a serious issue, and one not yet entirely reconciled, although great improvements have been made.

Bicycle transportation might have been a good topic for the authors, had they actually had a focus and real interest in transportation. Signal lights are an uncomfortable impediment that can make bicycle transportation less efficient and comfortable. Nevertheless, it seems likely there will be a convergence over time with progressive signal timing balancing out to 20 mph for dense urban areas. This is a speed that would allow most bicyclists to follow several blocks with the progression without stopping, while also serving the need to minimize the danger to pedestrians, since fatalities jump as speeds increase beyond this point. So despite some apparent drawbacks for bicyclists, the grid can also accommodate their needs rather well.

Perhaps if the authors were actually interested in transportation, when looking at the street grid, they might also have considered Portland's light-rail system - MAX. The length of MAX trains is limited by the short block lengths, and it would be tempting to identify this constraint as an argument for larger blocks. There would certainly be possible reductions in operating cost to add an additional car to the existing trains, rather than running more trains (labor costs for additional drivers, greater fuel efficiency, etc.). However, these operating cost reductions may not be optimal for system growth. The ability to increase service by increasing train capacity would likely slow down the addition of new routes, and the larger trains would lead to less frequent arrivals. These tradeoffs could ultimately result in a transit system with a network that is less dense and less convenient for passengers, which would ultimately attract less ridership. Obviously, a more detailed analysis would be necessary to understand both the short-term and long-term tradeoffs, but it seems likely the constraint imposed on MAX by the short downtown blocks is not to the detriment of developing the most effective transit system for the region.

Perhaps the authors could try again explain how the arithmetic could possibly work to increase on-street parking by removing streets. In the first instance, it appears they are talking about the number of spaces along a single street frontage, instead of inside the entire area. Sure, you can pick up a few spaces near the intersections by removing a street, but then you also lose two entire block faces.

Next, is sounds like they are retaining much of the street for use as a sort of parking lot. This is a minor variant on the grid at best, and one that would likely deteriorate the actual traffic operations in most cases if we were to model it, given the interruption of efficient street pairings and the additional idling and turning movements necessary to enter and exit these driveways.

This may be an admirable effort to apply general concepts and arguments developed in the abstract for other contexts, but there is no indication at all that there is any problem with the way Portland's grid operates. Sure, there are places along Burnside where the oddly shaped and poorly configured blocks create poor urban spaces. The capacity constraints imposed by the bridges could have been integrated into the street pattern better. The size of the blocks may, in fact, be shorter than optimal. Yet I find it hard to believe that the authors have actually experienced the streets in downtown Portland as dominated by "noise, exhaust, risk and the health and frustration of pedestrians" as they strain to suggest.

In my experience, downtown Portland is a wonderful place to be as a pedestrian! This is, naturally, in large part because I enjoy reaching the next block where I can take in another interesting architectural experience.

Portland's Grid

There are two important features of Portland's grid with small blocks that this post doesn't mention:

- Small blocks increase walkability. In fact, studies have shown that the frequency of intersections is the most important feature of urban design to get people to walk on more trips.

- Small blocks slow traffic and increase pedestrian safety, because drivers tend to slow as they approach intersections.

This post seems to be a grab-bag of criticisms of the Portland grid.

Some criticisms make sense:

- I think it is great idea to convert some street segments to parks, creating double size blocks. Pedestrians and bikes can still use them as streets, so walkability and frequent intersections remain. Paving is decreased, and more park space is added.

- It is useful to curve streets a bit in some places to eliminate endless vistas, as long as they are not curved so much that pedestrian routes become too long.

Other criticisms don't make sense:

- T-intersections improve traffic flow, as the article says, but they don't work well for pedestrians. When pedestrians get to the T, they have to go out of their way.

- Venice is indeed "the preeminent pedestrian haven." I hope that someday, we decide to build a sizable car-free city like Venice in the US. Until that day comes, Venice is not a relevant model, since we have to look for patterns that work for both traffic and pedestrians.

The criticisms are often inconsistent in their goals: eg, Venice is better than Portland because it is more of a haven for pedestrians (though it doesn't accommodate automobiles), and T-intersections are better than Portland because they improve traffic flow (though they don't work well for pedestrians).

Rather than this grab-bag of criticisms, it would be more useful to develop a consistent positive alternative to the Portland grid.

Charles Siegel

Favouring Pedestrians, Reducing Asphalt

Thank you for the points and the proactive take, Charles.
Here is stuff that I know, which reinforces some of your points:
Research does show a link between level of connectivity and walking activity. But connectivity should not imply asphalt; there are other options. More recent research reinforces your inclination toward less asphalt. Larry Frank and Chris Hawkins found that when we tip the balance of connectivity in favour of pedestrians, more walking occurs; paths and pedestrian streets do it. See their work here: Giving Pedestrians an Edge — Using Street Layout to Influence Transportation Choice

Stopping or slowing down at every 200 feet is very inefficient and highly polluting according to the EPA. It also creates driver frustration and antisocial behaviour with occasional dire consequences. For bicycles short blocks are a nightmare. And for pedestrians it means being on asphalt one quarter of the way anywhere, apart from the time lost waiting and the tension of dashing to make the light. For seniors and impaired people it can be an Odyssey.

Greening the streets is precisely the direction we aim for. Portland has done a bit and other cities more. See Vauban and Montpelier and of course Berkeley, CA and Seattle, WA and Vancouver BC were diverters and closures let pedestrians move freely but restrain the cars.

If properly planed in a new neighbourhood, T-intersections will find a corresponding path that continues a pedestrian’s route. We provided one solution to that question in the our previous OP-ED on Goodbee Square. Se the drawing in:

Venice is relevant as an ideal. This ideal has been driving changes in existing European city centres such as Munich, Essen, Montpellier and in new suburbs such as Vauban, Freiburg, now famous for its environmental stewardship.
You are right about moving on with designing the future. We produced at least one alternative (see Wiki: or Google: Fused Grid)

Can we talk about actual traffic engineering?

This discussion might prove far more productive if there were some grounding in basic traffic engineering. The statement: "Stopping or slowing down at every 200 feet is very inefficient and highly polluting according to the EPA" is inappropriate in reference to any modern grid pattern using progressive signal phasing. Traffic is enabled to continue uninterrupted, and, importantly, is directly discouraged from exceeding the speed limits. In these cases, traffic flow is very efficient for traffic and safer for pedestrians.

It also seems difficult to make a coherent argument that T-intersections are good for pedestrians, even if the direct-path issue can be resolved. Again, if we consider signal phasing, this becomes apparent. Because the T-intersections force vehicles to make (otherwise needless) turning movements, there are more vehicular movements that conflict with the pedestrians using the crosswalk. Although this problem could be mitigated by using a Barnes Dance, that solution can only be implemented by taking crossing time away from the pedestrians.

Perhaps the authors can address these issues of signal phasing.

Green Street Examples


Last note- I would like to see first hand or photos of the sorts of streescapes and street design you are supporting. The drawings don't mean much to me.

Are there any sites with photos?

Also, are there any examples in NY, NJ, PA or DE? I won't be making it to VT, OR or CA anytime soon.

Re smaller vs. larger blocks- in Center City Philly, the blocks are small. I agree that that is an issue re air pollution. However, it does seem to lend to closer knit neighborhoods, possibly. Out towards West Philly along some major city roads, the blocks are longer and take a lot longer to walk through- tiring to a pedestrian. They also have just as much asphalt as shorter blocks. I am not sure how block length influences greenness.


Many Examples

Europe is full of pedestrian-only streets and there are a number in the USA and Canada. In fact, in Europe there are whole downtown districts entirely free of car traffic. Supreme examples are Montpellier in France and Freiburg in Germany. Vauban, now famous for its no-car practices, can be seen in MS Virtual Earth in Bird's eye view. Search for "pedestrian streets" or "pedestrian precincts" or "car-free cities". Wikipedia has a list of the latter from all counties. I have some photos on my Flickr site. Search under images in google for Fused Grid. Another search title is "home zones" a program in England.

Improving the Grid

Fanis, your Wikepedia article about the fused grid is a valuable contribution to the discussion of how to improve the grid.

I would take issue with one aspect of your fused grid design: I think there should be connectivity across neighborhoods as well as within each neighborhood. In the illustration, your neighborhood streets seem to end at the major streets that divide the neighborhoods, but I think the street system should also encourage people to walk from one neighborhood to another: if you live near the edge of a neighborhood, you have many more destinations within walking distance of your home if you can walk easily to destinations in the adjacent neighborhood.

Let me kick in one more idea in response to your comment:

"Stopping or slowing down at every 200 feet is very inefficient and highly polluting according to the EPA. It also creates driver frustration and antisocial behaviour with occasional dire consequences. For bicycles short blocks are a nightmare. And for pedestrians it means being on asphalt one quarter of the way anywhere, apart from the time lost waiting and the tension of dashing to make the light. For seniors and impaired people it can be an Odyssey.

In Berkeley, CA, where I live, most of the city has a conventional nineteenth-century grid, with some blocks the same size as Portland's and some twice that size. The city has begun putting small roundabouts in many intersections in residential neighborhoods, with the goal of using roundabouts instead of stop signs. Short blocks with roundabouts at each intersection make traffic travel at a slow and rather steady pace; though they do slow down at the roundabout, there is not the stopping and starting that causes the environmental problems you mention. On longer blocks with roundabouts at corners, drivers speed up more between intersections. As a bicyclist, I like the short blocks with roundabouts, which make traffic much safer.

I agree with you that we should get rid of some of the asphalt and make some of those streets into parks with pedestrian and bicycle paths - keeping the connectivity for pedestrians and bicyclists but reclaiming land from the automobile. But in this case, it is still necessary to slow traffic at intersections of the ped/bike paths and streets to make the street system safe for all users. I think short blocks with small roundabouts at the intersections can do that - including the intersections of ped/bike paths and streets.

(Note: when I say short blocks here, I am thinking of blocks defined by ped/bike paths as well as by streets.)

Charles Siegel

Short blocks w/roundabouts


Do you know of any examples of this in practice on the East Coast?

Examples of Short Blocks With Roundabouts

I don't. You could try google. Or maybe someone else has examples.

Charles Siegel

Portland vs New York

Some clarification on Figure 1 would be appreciated.

It appears to suggest the short blocks in Manhattan are more than twice as long as Portland's blocks. This helps to create the impression that Portland's blocks are somehow extreme and unreasonably small. But the figure misrepresents the facts.

In fact, both cities have 200-foot long blocks (generally about 260 feet from center line to center line, although the major cross-town streets in Manhattan are about 40 feet wider). Anyone can feel free to check - for anyone without better tools available, Google Earth has a measurement feature that can give reasonable, rough estimates. Of course, the short blocks in Manhattan are generally recognized as a pedestrian-friendly size, whereas the long blocks have long been criticized.

It would seem rather difficult to look at Manhattan and conclude that short blocks have been a significant impediment to development. If they were so inefficient for developers, no business district ever would have grown so large!

The Commissioner's Plan

One can find the 1811 commissioner’s plan in the following links:

Best: the library of University of Texas historical maps where the comparative size of block is clearest: the old town, inherited, and the new city, planned:

Second best: Cornell Library:

2.Below is the image at:
In Wikipedia:
180 x 518 - 182k - png -
Below is the image at:

A partial modern drawing can be found here:

So, therefore....?

Can you please explain why your Figure 1 shows the New York short blocks as more than twice as long as the Portland blocks? As you recognize, the Portland blocks are 200 feet. Your reference number 2 clearly confirms the New York short blocks as "being in general about two hundred feet".

Why does the figure so grossly misrepresent both the reality on the ground and the original 1811 plans you reference here?

Another good source

There is also another good source that gives a good pictorial record of the growth of NY- "New York, the growth of the city" a small, illustrated book by M.J.Howard. On page35 it shows NY in 1755, more of a village than a city with its typical irregular blocks of all sizes. To judge and compare the size of these blocks one has to consider width and length which produces the block area. The same idea applies to the Commissioners Plan. It would seem though that these clarifications do not help answer the original question in the article. They simply assert that cities have blocks of many sizes.

(No subject)

Portland And New York Blocks

The labels of Figure 1 are a bit hard to understand, but it looks to me like the figure shows that New York's short blocks are the same length as Portland's blocks.

The blocks of Miletus are half the length of New York's short blocks. The problem is that the label for Portland is next to the line for Miletus. The label applies to the box it is in, not to the line it is near.

I don't think Fanis is saying that short blocks are an impediment to development. As I understand it, he is saying that the city would be more livable if we kept the connectivity that the grid with short blocks has for pedestrians and bicyclists, but we closed one-block stretches of some of the streets to cars.

From the wiki article about the Fused Grid: "A similar debate has also been taking place in Europe and particularly the UK, where the term filtered permeability [3] was coined to describe urban layouts which maximise ease of movement for pedestrians and cyclists, but seek to restrain it for motor vehicles."

Of course, Portland has closed two downtown streets that it has converted to transit malls. And I expect you could find neighborhood groups in Portland who would like to close one-block stretches of street to cars and convert them to parks with ped and bike routes. I think lots of people in Portland would like to keep the connectivity of the grid for peds and bikes but filter out some of the cars. You can also retrofit a grid to filter out through traffic simply by putting in diverters at intersections.

Charles Siegel

Good clarification

Thank you for the clarification Mr. Siegel. In my discussion with a friend from New York, we were obviously not attentive enough and simply followed the line across from New York and misunderstood the unusual alignment of the labels.

I appreciate the notion of creating functional differences within the street grid to serve different types of users. At the same time, I don't feel comfortable saying a street was "closed" in Portland for the transit mall. I would have to say the streets are still there, and are simply managed differently. The street grid is intact, even if the traffic patterns have changed.

Personally, I have always understood "street grid" and "blocks" to refer to public rights-of-way and development patterns. The streets in Portland were not de-mapped, and I would also contend that a pedestrian street is still a street. Given this understanding of the terminology, I am not finding the problems with the small-block street grid the authors try to suggest.

I would also note that the transit malls in Portland differ from the "Fused Grid" in a number of key regards. Perhaps most importantly, it maintains the paired one-way streets (and further demonstrated their utility by reusing one pair for a more intensive mode of transportation as the city grew larger!). I would also recognize the differences in pedestrian connections you identified. The "Fused Grid" looks very much like the suburban sprawl of Phoenix, and it does not seem to share much in common with the intensely active-yet-walkable urban districts that have proven popular and successful in North America.

I think it bears considering that using a fully-developed street grid provides greater flexibility for changes. While the transit mall in Portland has worked well, Eugene found its attempt at a pedestrian mall unsuccessful. With a full street grid, Eugene had the flexibility to redesign its traffic patterns and reconfigure its public space. The lack of flexibility inherent in the assumptions of a "Fused Grid" creates the obvious question about how well it could perform over the course of centuries as both transportation and land use demands change in unpredictable ways. The traditional grid already has a strong, proven record.

In terms of the street management approaches, I am not convinced closing one-block sections of streets is typically beneficial, as it forces more turning movements that conflict with pedestrians. Diverters often raise the same issue. Unless carefully studied from a network perspective, these local solutions have the potential to create problems for both pedestrians and vehicles (which include buses and delivery vehicles) elsewhere. Of course, depending on how special a particular location might be, the tradeoffs may be worthwhile. It is important to understand what the actual tradeoffs likely will be.

At the same time, a comprehensive study for pedestrianizing an area or diverting a major source of vehicles from a key location may result in greater gains for pedestrians. If a point further upstream in the network is identified, it may avert or mitigate the new turning conflicts, and extend the pedestrian benefits further than initially contemplated.

Fused Grid

"I would also contend that a pedestrian street is still a street. Given this understanding of the terminology, I am not finding the problems with the small-block street grid the authors try to suggest."

I agree. I like the idea of a fused grid with some pedestrian blocks that are closed to cars but are clearly part of a street grid that has as much connectivity as the traditional grid. Making the pedestrian streets look like streets makes the pathway clear to pedestrians and bicyclists. As you point out, it also allows flexibility to use the grid in different ways as the city grows.

Charles Siegel

Other Points about Portland's Grid

Some issues with Portland's grid beyond what people have mentioned is the small 200' x 200' grid does not allow for many street trees to be planted along a given block. With a 30' spacing and 35' feet planting from an intersection, a block will generate about 4 street trees being able to be planted. Portland's blocks can also limit the use of alley ways (you won't even have a backyard if you even tried) -- particularly in the more residential areas. The small sizes can limit the height of buildings, which can either be a good thing or bad thing.

And lastly, the small block sizes are not conducive to it's surface transit that runs along the streets. A light rail line can only carry two cars and often the trips from outside the downtown to the suburbs can be very crowded. Simply adding another car would alleviate this issue and add riders to the system, but it cannot as the LR cars would block the opposite intersection. The only option is to bury the transit lines which would be costly but very beneficial for numerous reasons.

So no, I do not think that Portland's small block sizes achieve anything profound. My personal favorite block sizes are at least 250' and no more than 600'.

MAX operations

These are good observations, and I don't really think 250 - 600 feet sounds unreasonable. I'm not particularly concerned about 200 feet, though. The effect of block length on MAX was actually brought up earlier in the discussions below. While it would certainly cost less for a transit operator to transport the same number of passengers on three-car trains than on two-car trains, it is not so clear that outcomes would be better if this option were available. It is an interesting and complicated question.

Burying the transit lines is not the only option; additional alignments can be taken through town. This has the additional benefit of denser transit connectivity.

First, consider that carrying the same number of passengers on three-car trains rather than two-car trains would result in a reduction in frequency of service. Reducing frequency of service makes transit less attractive, resulting in lower ridership. So again, the limitation on the length of the train may create conditions that attract more ridership. Conversely, it could result in fewer passengers per vehicle, but that is also less cost-effective, and takes away the pressure to expand the system.

So if you consider it from a game-theory perspective, it would be more likely that future MAX lines would not be put into service as long as there was idle capacity on the existing lines. Opponents would claim a new line was not necessary because the existing line "wasn't used," and there would also be internal cost-cutting pressures to make use of more feeder routes to make use of the existing rail capacity. Oddly enough, the more limited capacity imposed by the shorter trains may enable Tri-Met to add more routes to the system more quickly, and a more extensive network with more frequent service would attract more ridership than a more limited set of routes with higher capacity.

It is likely enough that the short blocks impose a somewhat higher operating costs, but may result in a better outcome.

The different types of systems - one with higher capacity routes concentrated on fewer corridors, or a more diffuse set of routes - would yield different types of metropolitan areas. The first would favor a highly peaked commute in and out of a strong CBD, while the other would be more ammenable to a multi-nodal metropolitan area. Many planners probably favor the latter, which is essentially the Portland model.

The idea of burying the transit lines raises a whole host of issues. Generally, I would not tend to favor it for Portland, because I think keeping the lines on the surface provides opportunities for traffic management and provides a superior passenger experience.

Of course, these are complex issues, and more study would be helpful to really understand all the tradeoffs involved.


I do not know where you are concluding that a three-car train would reduce frequency of service. There's no reason that a three car train couldn't be operated on the same headways as a two-car train.

Streetcars and buses, in my opinion, are a much better surface option for central cities than LR. Autos can run on the same lines and it does not restrict on street parking and bike lanes. Only LR cars can be used on LR lines, and I feel that is a major impediment to multi-modal transportation.

I have actually been on MAX trains that are too crowded to fit all passengers -- where some people are left behind. This is simply due to the fact that the streets will not allow for another car to be added. I'm very glad the system is popular, but for the future of the system it needs way more capacity and better times, all of which a subsurface system in downtown could achieve.

The issue still remains that a 200' block system is not conducive to LR transit beyond two cars. Good thing though, Portland can keep its existing surface lines in downtown and use them for streetcars.

Not impossible, but unlikely

There is no technical reason you could not run a three-car train as frequently as two-car trains. Basic political realities, however, make it unlikely it would actually be done. Transit systems, as government programs that rely on taxpayer subsidies, do not expand until they can demonstrate a need. With uncrowded three-car trains, there would be no need to run the trains as frequently as necessary with two-car trains.

Even though, as you observe, the Portland LRT gets crowded at peak times, there is still strong political opposition from many in the Portland area, and elsewhere within the state, that claims MAX runs "empty most of the time" and that "nobody uses it". Conversely, when people have the example of having to wait for one train it makes it much more difficult to successfully oppose expanding the system.

What I don't understand is the notion that a subsurface system is such a better choice. Since the system "needs way more capacity and better times" why wouldn't it be better to add additional routes, instead of spending all the capital funds to build a more limited set of routes with much higher capacity to connect very few places? It is worth noting that New York had a highly developed, saturated surface transit system before it became necessary to begin adding grade separated transit lines (and they did not have the benefits of modern traffic management technologies at their disposal). Portland may someday need such heavy-rail trunk lines, but it would benefit from much, much more network infill first.

The Portland grid is in no way abandoned

The authors approach this with a particular point of view and an unintegrated one where transportation, economics, planning and the environment are seen asstand alone parameters. Cities are not designed for efficiency but livability. Livability is first to be established and then priced. If efficiency were to reign supreme, Houston and Irvine would be the hottest cities in the country. It is also unfair to compare a rectangular grid with an angular one- they are different flavors. Lastly, I dont have anything against 3-way intersections but one way streets are not the best solution for city streets. One way traffic is not good for retail- very few suvccessful main streets are one way streets. One reason is that access to stores is more difficult and these streets encourage faster movement of cars.

Within rectangular grids, one may argue about the final block size but Portland is as as small as downtown block sizes get. It is a pleasure to walk in the city. And the city has been very creative in making floorplates and parking work at that size. The size of the blocks ascertains how much the public realm wraps around and infiltrates the private realm. In that vein - the smaller block size the better as that leads to a more vibrant public realm and an accessible and walkable city.

The Portland grid is in no way abandoned- it has inspired an awareness and control of block size and a move away from superblocks. The biggest criticism is that that to satisfy the parking demand of a block, underground parking is the way to go, which is cost prohibitive. That it is hard to achieve the overground donut parking model (uses on street face, parking in the middle), which is the cheapest hybrid of cost meets urbanity. These blocks are usually a minimum of 250' x 350'. But that is a very temporal problem and can be solved with district parking strategies in the short term.

Much better than walking endlesslessly along a block, waiting for the next traffic light in Irvine....

sameer chadha

Angular grids etc.

What are people's favorite angular grid- based cities?

Can you give examples of successful Main Streets with two way streets?

Regarding community building, while I recognize that dense neighborhoods lends to greater community, I think it's all about effort. Real community takes effort- resident meet-ups, cultural events, parties, dinners etc. I have seen some suburban neighborhoods with more community than some urban neighborhoods.


In response to "Irvine would be the hottest cities in the country"

Irvine is the most expensive OC inland city, even by SoCal standards it's expensive, certainly by Portland standards. It's an inconvenient truth for us walking advocates, I wish this wasn't so, but many people love Irvine, and their cars. Also Irvine has extensive park/bike path network with over 70miles of dedicated bikepaths for recreational riding.

What we need is true urban area with low crime, good schools and walkability, doesn't exist in US yet, even in Portland suburban schools are better and crime lower.

In defense of the standard grid: it's about legibility

Please see my critique of this article here:


Jarrett Walker

Prepare for the AICP Exam

Join the thousands of students who have utilized the Planetizen AICP* Exam Preparation Class to prepare for the American Planning Association's AICP* exam.
Starting at $199
Planetizen Courses image ad

Planetizen Courses

Advance your career with subscription-based online courses tailored to the urban planning professional.
Starting at $14.95 a month
Book cover of Essential Readings in Urban Planning

Essential Readings in Urban Planning

Planning on taking the AICP* Exam? Register for Planetizen's AICP * Exam Preparation Course to save $50.
$369 includes free shipping!

A Short History of America

From comic book artist Robert Crumb, poster shows how the built environment has changed throughout the decades.