Penn’s Jacobsian experience, and the difficulty of planning

The frontispiece quote from The University and Urban Revival: Out of the Ivory Tower and Into the Streets, by Judith Rodin, president of the Rockefeller Foundation and former president of the University of Pennsylvania, comes from Jane Jacobs's’ The Death and Life of Great American Cities:
No single element in a city is, in truth, the kingpin or the key. The mixture itself is kingpin, and its mutual support is in the order. A city’s very structure consists of a mixture of uses, and we get closest to its structural secrets when we deal with the conditions that generate diversity.

Planning vs. Big Plans

And Penn’s path to local redevelopment, after decades as the eight-hundred pound gorilla, depended on planning, but not on monolithic Plans. Indeed, Rodin warned at a recent panel that Penn’s experience was its own, not a blueprint. Still, a caution is valuable.

She writes that Penn decided to roll out implementation of the West Philadelphia Initiatives rather than announce a comprehensive master plan:
In the matter of testing and learning, our view was based on the work of Charles Lindblom and James Scott, each of whom studied failed attempts at social change. In his 1990 book Inquiry and Change: The Troubled Attempt to Understand and Shape Society, Lindblom points out the impossibility of being truly comprehensive in urban planning from the start, because there are inevitable biases that frame the work in the abstract and there is an inexhaustible number of forces that enter into the life of cities over time that cannot be anticipated in advance. Among the recommendations to social change that Scott [in Seeing Like a State] advocates is to take small steps in an experimental approach, with the presumption that one cannot know the consequences of every intervention in advance.

This raises many questions about the environmental review process regarding major projects like Atlantic Yards. Among them: Can a ten-year effect on traffic and transit really be estimated? Is ten years a legitimate endpoint, or an arbitrary one? And what if the buildout would take much longer?