To Achieve Sustainability, We Need to Go Upstream in the Design Process

Over the last few weeks, as we’ve all crammed more and more parts of our lives into our homes, many things have changed. One significant shift is that our overall energy use is down. Unsurprising perhaps, with travel on hold, more people working from home, and bars and restaurants shuttered. More of us at home, though, means residential energy use has spiked.

Yet, who could blame people for using more energy at home, while wisely following public health guidelines on social distancing?

This dramatic shift in energy use brings an important aspect of behavioral design into focus.

Let’s rewind one year, pre-pandemic.

It’s April 2019. If you wanted to reduce the amount of energy people were using at home, how would you have used behavioral science to do it? Would you have tried to make energy and cost savings more salient for residents through an electronic dashboard? Perhaps you’d have advocated for comparisons and competitions between neighbors. Or maybe you’d have encouraged people to install a smart thermostat and monitor their energy use throughout the day.

Behavioral science is a powerful tool to direct human behaviors toward sustainable outcomes—but too often, behavioral science has focused its efforts on changing the behavior of end users of a system. Rather than focusing on how to change the behavior of end users, a more effective energy-saving intervention might be to shift the focus to earlier in the design process—to those who designed the building, the local energy grid, or renewable buyback policies.

Consider how our home energy use might be less today had behavioral scientists intervened further upstream in the system.

Exploring the untapped potential of design behavior for sustainability is the focus of a recent expert panel report from Nature Sustainability. The panel, organized in part by my colleagues at the Convergent Behavioral Science Initiative at the University of Virginia, brought together experts in academia, practice, and policy for two days of in-person discussion. Over a subsequent year of virtual collaboration, 20 key questions about behavioral design for sustainability emerged.

Why is design so important for sustainability? As the report argues, while “end-use behavior can determine what happens in a situation, design behavior often determines the situation itself.”

While end-use behavior can determine what happens in a situation, design behavior often determines the situation itself.

To learn more about how the panel and report are catalyzing new ways of thinking about and applying behavioral science, I reached out to five panelists to pick a key question from the report that is changing how they engage with their work. Below, they explore ways we might peek behind the curtain to better understand (and influence) how designers make design decisions, why engaging with diverse stakeholders is critical, the importance of timescales, and more. Read on to see the questions they selected and how answering them might help us address urgent sustainability challenges.

What timescales do (and should) designers consider?
Patrick Hancock, Ph.D. fellow at University of Virginia

What we design today reaches far into the future, likely even beyond designers’ expectations. For example, a bridge you used to get to work this morning might be used by your grandchildren to commute to their jobs. Let’s take a look at the case of the Brooklyn Bridge, which has been around since 1883. For a bridge to survive so long, the original design has to be both resilient and adaptable. Initially created for horse drawn carriages, trolley cars, and pedestrian foot traffic, it now has six lanes for cars. In the case of the Brooklyn Bridge, uncertainty surrounding design safety, such as ensuring the bridge stays standing under expected loads, partially explains its resilience. With that said, what a designer thinks is a relevant timescale might not always match the longevity of the things they create and the needs of the people who use them in the future.

Growing evidence suggests perceived timescales can have profound implications on design outcomes. For example, Christina Atance and Kevin O’Neill’s theory on episodic future thinking posits that individuals’ decision-making can be improved by “pre-experiencing” decisions before they occur. In a sustainability context, this might look like the modular design of the California high-speed rail bridges in which designers’ anticipation of possible climate futures facilitates a more adaptable—and therefore resilient—design.

How does framing the impact of a designer’s decision far into the future, say 100 years versus 15, impact the quality of their sustainability-relevant decisions?

To try to understand how timescales can help motivate climate action, my collaborators and I explore the impact of future visioning (akin to episodic future thinking) on how decision makers design. For example, how does framing the impact of a designer’s decision far into the future, say 100 years versus 15, impact the quality of their sustainability-relevant decisions? By better understanding the timescales designers have in mind and how that impacts their decision-making, we can hopefully design targeted interventions which improve the sustainability of their designs.

With this in mind, we can hypothesize that imagining the future could improve the present-tense design process for more sustainable outcomes. Although there will always be uncertainty around how infrastructure will adapt to physical changes in the environment, there is also a great deal of uncertainty regarding how designers make decisions. In order to unlock design improvements, we need to focus our efforts on understanding the latter.

How can designers engage with communities and users?
Ruth Schmidt, associate professor at the Institute of Design, Illinois Institute of Technology

To improve design decision-making, designers must engage with end users to design with their values and needs in mind. Before we decide how to engage end users, however, we must first unpack who qualifies as a user, especially in complex systems where we are likely to juggle design considerations for many different constituencies. If we were to add a greenway with bike lanes to a city block, for example, bikers would obviously be our primary audience. But designing for the needs of this one group (who might prioritize the expanding the width of the bike lane in the interest of safety) without also considering the needs of others (such as shop owners or residents who need parking access) might result in an overall reduced quality of urban life. We benefit from engaging diverse viewpoints in all sustainability design contexts. Not only does this give us insight into stakeholders’ differing mental models and needs, but it helps us understand and anticipate how those needs might intersect or be in conflict.

We can use what we learn to design solutions that are more proactive than reactive, preventive and responsive.

A deeper understanding of these interconnected user contexts can also help us design with a more longitudinal view: what emergent needs and behaviors must we consider? Who should “own” solutions, to help adapt them over time? Who needs to be included if solutions are to be perceived as legitimate in the first place?

Once we’ve developed hypotheses, we can also engage users in prototyping potential solutions. This helps to identify what’s not working or where solving one set of problems might contribute to downstream issues earlier in the design process, while there is still time—and budget—to course correct. Prototyping is a complement, not a substitute, for randomized controlled trials, more an opportunity to refine hypotheses than to confirm them. But when done well, we can use what we learn to design solutions that are more proactive than reactive, preventive and responsive.

How can this community of scientists influence policy?
Karen Akerlof, assistant professor at George Mason University

Some of the biggest barriers to the sustainable design of buildings, infrastructure, transportation and energy systems, and green corridors are outdated public policies. Layers of local, state, and federal laws that were enacted decades ago—before sea level rise threatened coastal cities, transportation became the biggest source of U.S. greenhouse gas emissions, and species habitats started to shift with changing temperatures—now either raise legal questions or serve as outright impediments when designers try to address climate through mitigation and adaptation.

Layers of local, state, and federal laws that were enacted decades ago—before issues like sea level rise threatened coastal cities—now either raise legal questions or serve as outright impediments.

For example, states that apply Dillon’s Rule—a legal principle that restricts the authority of local governments if they lack official sanction by the state—can place limitations on their ability to make on-the-ground decisions in these policy areas. Local laws themselves may also serve as a barrier. In order to promote more sustainable decisions about building in coastal areas, for example, one of the first considerations may be changes to regulations, such as zoning laws and increased “set back” distances to discourage construction in areas likely to be affected.

Engaging with policymakers and stakeholders to understand where the blocks and bottlenecks occur—and in what instances community values and political will support change—is an important role for social and behavioral scientists. Designers of sustainable systems must become savvy about understanding—and shaping—policy in order to implement their work.

How can researchers and practitioners work together?
Baruch Fischhoff, Howard Heinz University Professor at Carnegie Mellon University

People who study decision-making evaluate fixed options. With sustainability, those options might be specific “green” products, socially responsible investments, or cap-and-trade programs. We have little to say about the source of those options. That focus on fixed options undermines the creativity needed to create new ones, while letting the research be driven by the agendas of those whose options are being analyzed.

In contrast, designers create options, uncertain how they will be evaluated. They often have direct contact with people affected by those options, helping them to use that creative freedom. However, their work is often guided by their intuitive theories of human behavior, rather than by behavioral research that might help them to overcome the natural limits to understanding people with different backgrounds, perceptions, needs, and aspirations.

Researchers and designers would do well to work together, learning from one another. That will require bringing them together throughout the design process, from project conception to execution and monitoring. For behavioral researchers, such collaborations will require professional rewards for the kind of “slow research” that participating in such creative processes entails. For designers, they will require patience in connecting research abstractions with practical applications.

How can evaluation balance perfection and relevance?
Erin Sherman, vice president at ideas42

The perfection versus relevance quandary refers to the trade-off between only sharing research findings that meet traditional standards of rigor and the urgent need to adapt and scale sustainability interventions, even if they aren’t “perfect” by those standards. I think of this quandary as linked to the questions: What knowledge gets disseminated and used? By whom? And how?

I’ve found Mary Ann Bates and Rachel Glennerster’s piece on the generalizability puzzle to be helpful here. To ground this quandary in an example, consider transportation behavior. In 2019, my team at ideas42 started working regularly with transportation demand management professionals through the American Cities Climate Challenge. They have been working to reduce urban congestion and encourage alternate modes of transportation, from biking to carpooling to busing, for over forty years. In that span of time, they have undoubtedly accumulated expertise, noticed patterns, and shifted how they make decisions––but as we searched for peer-reviewed literature about transportation behavior change, there were precious few examples.

Evaluating transportation demand management programming is expensive and randomization is rarely feasible. When your employer has to stretch and scrounge to find funding for pre-/post-measures of behavior, it makes sense that your standard of proof for an effective intervention would quickly run into a ceiling. It would be an even further stretch, with even less certain benefits, to attempt to collect data to carefully and consistently compare the context of your intervention to that of other cities or towns.

But a dearth of academic literature does not mean that dissemination has not happened; it has just happened differently. The transportation professionals and other career practitioners disseminate understanding through sharing stories and comparing notes on their contextual challenges and solutions. Of course, while storytelling is immediately rewarding and engrossing, is hard to scale and prone to error. As a behavioral scientist, I am curious if we can harness what we understand about the hedonic and social value of storytelling to make dissemination both more systematic and more likely to take place.

I’m encouraged by the development of networks like the panel of contributors to “Twenty Questions” and the American Cities Climate Challenge transportation behavior cohort. If nothing else, these peer-to-peer networks will accelerate dissemination through storytelling, which can help prompt progress. At best, they will help develop new language, norms, and reputational benefits for systematically evaluating context along with behavior.

These reflections represent just a small portion of the report. The full report with all twenty questions about design behavior for sustainability can be found here. Some of the questions that I find compelling that we didn’t have a chance to discuss here include probes into how design behavior is different from end-user behavior, as well as questions around how singular design decisions ripple out to impact and influence others.

These questions are an important but preliminary step in understanding and leveraging design behavior for sustainability. Answering them to create actionable solutions to climate change will require rigorous collaboration from people across disciplines and fields. If you’re a behavioral scientist, consider reaching out to industry professionals as you shape your next project. What are the behavior change questions they grapple with? What mental models do they bring to the design and implementation process? Likewise, if you work in industry or policy, it would be worthwhile to ask how you can incorporate behavioral science in your next endeavor. Is it possible to add more measurement and rigor? How can we design to ensure maximum learning? Behavioral scientists can help answer these questions.

For all of us, rather than thinking about how we might change the behavior of end users, consider how we can change behavior further upstream for greater impact—even turning the lens of behavior change to ourselves. Whether you design studies, infrastructure, or policies, rather than focusing on how downstream users might change, think about altering your own behavior to create preferred sustainability situations for them instead.