Solyndra, Moneyball, and Lessons for Planning
The Los Angeles Times recently had a story about the collapse of Solyndra – the once heralded poster-child of the Obama administration’s green jobs plan. A big part of Solyndra’s demise was due to the rapidly falling price of their competitors’ solar panels. In 2008, the cost of solar panels was a bit over $4 for each watt generated. Solyndr
The Los Angeles Times recently had a story about the collapse of Solyndra – the once heralded poster-child of the Obama administration's green jobs plan. A big part of Solyndra's demise was due to the rapidly falling price of their competitors' solar panels. In 2008, the cost of solar panels was a bit over $4 for each watt generated. Solyndra could beat that, offering a watt of solar power for $3 per panel. Then the Great Recession hit, energy demand fell through the floor, the price of silicon plummeted, and China aggressively cut the price of their solar panels to maintain market share. Prices dropped as low as $1.25 per watt, while Solyndra's price remained at $3 per watt. Game over.
The Solyndra bankruptcy gives us a window to understand a common planning shortcoming. Planners like visionary ideas, but are too often undisciplined in applying systematic data and principles to their search for the "next big thing." Green jobs sound good, and the L.A. Times article noted that by 2009 lots of smart money had already invested in Solyndra. Yet the mistake, clearly foreseeable at the beginning of 2009, was that even the best alternative energy company would have a hard time succeeding in the face of rapidly dropping fossil fuel prices. Let's move from the debate about Solyndra to the broader lesson for planners.
The common planning approach, of instinct, fashionable ideas, and a dash of "herd behavior", hardly differs from the old-school approach to evaluating baseball players. In Michael Lewis' Moneyball, and the Brad Pitt movie, the old-school scouts used their instincts to evaluate talent. Sometimes they called it right, sometimes not, and some scouts were better than others, but it was an "all art, no science" approach to the game. The Oakland A's won over 100 games in 2002 by applying statistical analysis to more reliably judge talent. Oakland still took chances, but they endeavored to build a fact-based understanding of the gambles they were taking. We in planning are too often the "gut instinct" scouts who put too much faith in our experience and too little emphasis on systematic principles and analytics.
In some of our more design-oriented planning schools, context and case are emphasized. Those schools turn out planners that are more art than science. Sometimes they guess right, sometimes not, but there is not enough systematic analysis and data collection behind their decisions to build an influential field. On the other hand, a smaller group of planners has venerated market-oriented mechanisms to the point that they have lost their view of context. My point is not that either design or economics is superior to or subservient to the other – certainly that one-dimensional debate has gotten us nowhere. My point is that planning should aggressively nurture the combination of the systematic and the contextual, and that we should excel in such a hybridized approach to policy analysis.
The land use – transportation debate illuminates this point. Some planners argue for transit-oriented development, while other planners argue for efficient pricing as if pricing alone were a panacea. I have written elsewhere about the limited view of both schools, and about how a more powerful approach would unite market principals and transportation pricing with land use and place-making approaches. The two approaches – urban design and pricing – are complements, not substitutes.
For such a hybrid approach, what is our planning equivalent of sabermetrics? We need data on interactions. How much do persons drive when faced with different congestion or externality prices? How much does transit ridership increase as parking prices are changed to tilt the field away from solo driving? How might carbon taxes (or cap-and-trade permit prices) influence the market penetration of electric vehicles? How might the effect of electricity price on electric vehicle adoption vary in neighborhoods with different densities, land use mix patterns, and demographics? We do not have nearly enough information on such questions, nor do we train our students to think in this manner.
Planning traditionally has occupied a middle ground, between abstract policy analysis that was too blissfully ignorant of place or context and case analyses that lacked sufficient grounding in theory. That middle ground, in the 1960s, spawned a nascent effort at developing systems thinking within planning. Partly because of lack of computing power and data, and partly because the thinking at that time was too focused on programming and not sufficiently aimed at training professionals with tools to think systematically about the interaction between abstract principle and specific context, the effort foundered. Planning should revive the effort. We are currently little more than the old, graying baseball scouts of the pre-Moneyball era, trading on our instinct but offering little in the way of broader systematic understanding. We can do more, and the sustainability challenges of our era require that we do more.