The Rush to Resilience: 'We Don't Have Decades Before the Next Sandy'
In an era of disruptions so significant that we refer to them in a single-name shorthand (think: 9/11, Katrina, Fukushima, Haiti, Sandy) what gives communities their ability to bounce back? And what does it mean for the way we build (and rebuild) cities? To explore these topics, we invited two leading thinkers who are working at the forefront of resilience — Andrew Zolli and Jonathan Rose — into a dialogue.
Zolli’s new book, Resilience: Why Things Bounce Back, written with Ann Marie Healy, is a must-read primer on the field of resilience research, examining the capacity of everything from people, organizations, communities, and societies to adapt to volatile and dramatically changing circumstance. Zolli’s day job is running PopTech, a network of cutting-edge scientists, technologists, and social innovators who come together to work on new approaches to some of the world’s toughest challenges.
Jonathan Rose is an urban planner, developer of green affordable housing, and the founder of the Jonathan Rose Companies. His work is focused on the intersection of climate, cognition and behavior, having incubated a pioneering network that explores these topics at the Garrison Institute.
Here’s part of our wide-ranging conversation about how to make our cities more resilient.
RF: Let’s get things started with the basics. Resilience is becoming a theme among city-builders and in urban planning circles. How do you define resilience?
AZ: Resilience can feel frustratingly slippery to define, in part because different fields use the term to mean slightly different things. In engineering, for example, resilience generally refers to the degree to which a structure like a bridge or a building can return to a baseline state after being disturbed. In emergency response, it suggests the speed with which critical systems can be restored after an earthquake or a flood. In ecology, it connotes an ecosystem’s ability to keep from being irrevocably degraded. In psychology, it signifies the capacity of an individual to deal effectively with potentially traumatic events. In business it’s often used to mean putting in place backups (of data and resources) to ensure continuous operation in the face of natural or man-made disaster. Though different in emphasis, each of these definitions rests on one of two essential aspects of resilience: continuity and recovery in the face of rapid change. And, in our volatile age, they’re all going to be part of the field we call urban resilience.
A good working definition, particularly in an urban planning context, is: the ability to maintain core purpose, with integrity, under the widest variety of circumstances. More broadly, it’s the ability to recover, persist or even thrive amid disruption.
Among other things, resilient systems sense and respond to their own state and the state of the world around them, compensate or dynamically reorganize themselves in the face of novel shocks, decouple themselves from other fragile systems when necessary, fail gracefully, and have strong local self-sufficiency.
But we shouldn’t get too hung up on the specific language. Any definition will only ever be a first approximation, because there are countless kinds of systems, countless ways for them to be more or less resilient, and countless things for them to be resilient to.
RF: Tell us more about urban infrastructure specifically. How can we make it more resilient?
JR: That’s particularly true in the world of urban infrastructure, which is actually comprised of many different kinds of systems. These can be typified as hard or soft, central or distributed, passive or energized, complicated or complex, synthetic or natural. And the more diversity they have – providing different but redundant pathways – the better. Urban infrastructure systems actually call on all four of the kinds of resilience that Andrew mentioned — engineering, ecological, business, or operational, and emergency.
For example, New York City's water supply and distribution system is natural, hard, passive and complicated, and, this makes it pretty resilient.
It’s ‘natural’ at its supply end – the water comes from rain, lakes and reservoirs. And when the city was faced with a mandate to reduce the quality risk to its supply, it also chose a ‘natural’ solution – to buy up the land around the reservoirs rather then to build a much more sophisticated and complex electric powered water filtration plant. The natural system is also passive - so it works whether or not there is electricity. And the water flows from the reservoirs to the city by gravity, passively rather then through an energized system. If the electric power had gone out at a filtration plant after Hurricane Sandy, the city would have been without clean water.
It's also a hard, complicated system, meaning that it's linear – just pipes and valves, versus a complex system, which has infinite interdependencies. Complex systems are much more tunable, and therefore more responsive to subtle variations, but also much more likely to malfunction in situations of macro variation. And finally, its fairly centralized-which means that if one of the two main water tunnels or control systems should fail, the whole system fails. The city is building a third water tunnel – which increases the diversity of the system by 50 percent.
So the first step in urban resilience is to make sure that all of our hard infrastructure systems can operate as passively, and linearly as possible in an emergency mode. We should use all the advantage of complexity and software to make them more efficient and responsive when systems aren't failing, but they should have a default mode that is passive, linear and functional.
AZ: Jonathan’s example of the New York water system points out the difficulty and nuance of assigning the term ‘resilient’ to any piece of urban infrastructure. There’s a paradox here. On the one hand, aspects of New York’s water system could rightly be called fragile. By modern standards, some of it is ancient – it still has wooden elements put in place 100 years ago – and, its centralization in places creates potential bottlenecks that might diminish its resilience in a crisis. Also, it’s not exactly self-reconfigurable – knock it over and it won’t put itself together again.
On the other hand, it’s also ‘dumb’ in the passive (and positive) sense that Jonathan mentions, and therefore avoids the much-more-common, contemporary ‘complexity risk’ that can come from having systems that are laden with complexly-interacting control systems, like those found inside a modern jumbo jet. There, when things go wrong, you might be able to hit a series of buttons and recover. However, on a clear blue day, you could also accidentally spill your coffee, push the ‘wrong’ button and crash the plane - which is why we so often hear about ‘pilot error’ causing crashes. You just don’t hear those kinds of story about the water system. There’s fewer buttons for Homer Simpson to push.
The larger point is that there’s always tradeoffs in these kinds of systems – tradeoffs between their complexity, controllability, efficiency, redundancy power, cost, riskiness, responsiveness, resilience … you can’t have it all.
So the first question is really, how do you make the best tradeoffs within the constraints you’re facing, for the times you’re living in? We built our infrastructure to work within one set of constraints, but those have changed, and we will need to change with them. But that’s not how most people think. It was just hours after Hurricane Sandy made landfall that people started asking “do we need to put a wall around Manhattan?” It was like a psychic rejection of the whole phenomenon.
But let’s say you were able to do that: wall the whole city off. You know what you’d get? You’d get Category 3 walls followed, eventually, by Category 4 wind and water, and then utter devastation. Then you’d rebuild Category 4 walls, and everything would be fine for a while, until you’re on the receiving end of Category 5 wind and water, and then, again, disaster. You cannot harden the shell tough enough for the egg not to crack eventually – especially if you otherwise follow business as usual.
The No. 4 reactor building of Tokyo Electric Power Co. (TEPCO)'s tsunami-crippled Fukushima Daiichi nuclear power plant on Feb. 20, 2012, nearly a year after the tsunami and earthquake which triggered the world's worst nuclear crisis since Chernobyl. (Issei Kato/Reuters)
RF: Some people say let's just do a better job of protecting cities - we can solve the problem of coastal flooding, for example, by building bigger flood walls. What does resilience do that prevention doesn't?
JR: One of the problems with more walls is that it doesn’t address the issue of more water – it just moves it elsewhere. If we could really wall off the water from lower flood zones, it would just push it to the next zone.
Civilization emerged in the flood zones of river valleys, where the soil was the most fertile. These are the highest risk zones, but also the most productive. Today, eighty percent of the world's economy takes place along side of rivers or waterfronts, so the entire global economic system is at risk of sea level rise and increased weather volatility. (And much of the rest of the world is at risk of droughts.) In every case, humans took the soft, natural and variable edge of a water/ land system and made it a hard edge. The world is facing a huge cost if its solution is to build ever-increasing walls. And if one fails, it’s a total failure for the city.
One of the lessons of Hurricane Katrina was how valuable natural wetland edges are for absorbing the water, surge and energy of a storm. We will probably have to install more hard systems such as sea gates at the entrance to our city's harbors. But we also have to rebuild natural wetland systems, or hybrid natural/ constructed systems that can absorb nature’s weather variability.
And we have to wake up to the issue that our cities and countries simply cannot afford the next century of climate risk in many of our current locations – we are going to have to start a movement upland or to higher elevations. For example, the elevations of the Metro-North train lines that travel up the Hudson and Long Island Sound are going to have to be raised – and that going to take decades.
AZ: I can imagine a lot of municipal managers and policymakers doing a spit-take reading those last couple of sentences. Adding redundancy and diversity to infrastructure is expensive enough, never mind the cost of raising the rail lines – and many of these regions are deep in the red as it is.
In many ways, this situation – upgrading infrastructure that’s heavy, expensive, rigid, poorly placed for the times, hard to move, and (lest we forget) mired in bureaucratic red tape – is akin to the situation in many parts of the developing world. (Indeed, climate change might be the great leveler in this regard.)
But therein may lay interesting ‘leapfrogging’ opportunities. In India, for example, in early the 1990’s, the waiting time for a landline telephone was more than a decade. (Imagine that you moved into a new home on the day Bush beat Kerry – you’d still be waiting for a phone line until at least 2014!) Back then, there were very smart people trying to figure out how to cut the wait time in half – to just five years. And then of course the mobile phone came along and made the whole exercise moot.
So in some cases perhaps we shouldn’t be thinking about moving pieces of old infrastructure, but abandoning them for something else entirely – stuff that’s lighter weight, lower-density, cheap, movable, disposable, redundant and designed to fail – in other words, re-balancing our design practices within a different set of constraints. This is, for example, the way that Google server farms are built – individual parts fail all the time, and are continuously replaced.
But whether we’re looking at decades of re-elevating the trains, or decades of re-inventing the next great replacement for our current transit systems, it doesn’t matter in the short term, because at the rate we’re going, we don’t have decades before the next Sandy. And that makes a second question even more important, which is: how do you enable the right kinds of improvisational, creative responses when things inevitably go wrong? That, in turn, takes the conversation out of the world of ‘hard’ infrastructure and into the realm of ‘soft’ infrastructure.
A flooded intersection in Belmar, New Jersey, on Nov. 2 in the aftermath of Hurricane Sandy. (Tom Mihalek/Reuters)
RF: Now that's interesting, these two dimensions of infrastructure - "hard" vs. "soft." Tell us more about that?
JR: Soft infrastructure are our operating systems: for example, decision making, governance, finance and management systems are all soft infrastructure systems, and these need to be designed to work better in emergencies.
That, in turn, means they must fulfill two key, yet opposite requirements: first, they really need to be robust, and have independent redundancy (for example, servers doing the same thing on two totally different grids). Second, the people and systems who are being connected need to know what to do if they become disconnected - they need to be able to act with full authority independently if need be. To do so takes training, and a wide range of cognitive skills. Emergencies can be very traumatic – one of the clear issues that happened during Hurricane Katrina was that so much of the city's leadership was personally traumatized from their own losses that they lost the cognitive resilience to lead.
This need to operate both together and apart is mirrored in non-human systems as well. We need systems that are integrated across scales, but also separable across scales. For example, Con Edison currently requires any solar, fuel cell generator, or other independent power system on the grid to disconnect when the system fails. In fact, we need to empower these systems to to just the opposite – to kick in when the larger systems fails. Just the way that we need to be able to integrate centralized and distributed leadership, we need to do the same with things like energy and data.
RF: In your book, Andrew, you argue that resilience is frequently more about enabling people, neighborhoods and communities to respond faster and adapt better than it is about systems. That seems like a really important distinction.
AZ: Note the paradox here. Jonathan is exactly right: there are plenty of systems that achieve resilience by embracing modularity, interoperability, and a kind of distributed localism – things like electrical microgrids and wireless mesh networks are fine examples. And yet, earlier, we also discussed aspects of the resilience of the New York water supply, a big, passive system. The point, again, is that there are lots of different pathways toward resilience. What you need, as we mentioned earlier, is diversity – of scales, and of modes of operation. In other words, you need both/and, not either/or.
But to return to the human dimension: the thing about all disruptions involving people, particularly in cities, is that each one is different – and so, while rehearsal and preparedness is hugely helpful, you can never be perfectly prepared. Resilient responses therefore tend to be ‘adhocratic’ – mixing both a formal institutional response and informal, often citizen-led response in an improvisational stew. It’s more like a game of pick-up basketball than like a symphony orchestra.
One way to make cities more resilient, therefore, is to design in more opportunities for these improvised responses to occur. For example, in the wake of the Haitian earthquake, a global group of technologists build and deployed a system in 96 hours that allowed ‘crisis mappers’ around the world to map damage and urgent needs in Port-au-Prince, translating texts from citizens who were still trapped in the rubble. Another group, called Means of Exchange, is enabling alternative forms of economic exchange, which can be vital when there’s a shock to the money supply, as there often is after a natural disaster and there’s no power to run the ATMs. Yet another group, called Open311, is working on extending city information services.
All of these efforts help people self-organize, build greater local self-reliance and care for one another more than they ‘have’ to – and that’s the very definition of a resilient community.
Yet, today, most of our focus is solely on how governments and NGOs respond to disruptions – everyone else is a ‘consumer’ of their response. We bias in the direction of the formal bureaucracy, and treat everyone else as victims or onlookers - but there’s so much capacity there, in the heads of those citizens!
And we’re going to have to pay a lot more attention to those heads for another reason too: just as the costs of the disruptions goes up, so too will the costs of human stress – on both the physical and mental health of people experiencing disruption.
JR: The good news is that over the last ten years there has been an incredible flourishing of research on mental resilience. Just as we need flexible, adaptable, responsive infrastructure systems, we also need flexible, adaptable, responsive cognitive systems when crisis comes. And when a crisis comes, our natural wiring takes us right to the amygdala, the flight/fight part of the brain. That puts us into a single issue/single response mental mode exactly at the time in which we need to be evaluating multiple solutions and responses.
At the Garrison Institute, Diana Rose and Sharon Salzberg have been focusing on how to help generate this flexible cognitive capacity in the front line emergency responders and the aid workers who deal with the overwhelming trauma of extraordinary crisis’s. These workers throw themselves fully into the emergency, and rapidly burn out from stress, and from secondary trauma, literally taking on the trauma of the people that they are dealing with. This significantly reduces their capacity to be effective, to make smart, on the ground decisions. We saw this in Katrina, and recently in Haiti, and it’s a problem in every emergency room and refugee camp around the world. Interestingly, the latest neuro- and cognitive science are pointing towards the oldest mind sciences as part of the solution. Meditation, it turns out, is one the most efficacious ways to build mental resilience, and to train the people to stay in a integrative, problem solving state of mind rather then a re-active, overwhelmed state of mind.
AZ: So it’s clear we actually need resilience at multiple levels – from the psyches of individuals to communities to large-scale systems. We’re just at the beginning of understanding how these different pieces fit together, and how efforts at different scales interact with each other. While the fields aren’t new, the synthesis is – and in a world where volatility is the new normal, it couldn’t come soon enough.
RF: Thank you both very much.
Top image: A man walks through piles of debris outside of homes damaged during hurricane Sandy in the Queens borough neighborhood of Belle Harbor, New York, on Nov. 8. (Lucas Jackson/Reuters)