Institutional insights on integrating social and environmental science for solutions-driven research

Abstract

Solving complex environmental problems requires interdisciplinary research involving the social and environmental sciences. The U.S. Environmental Protection Agency is working toward solutions-driven research and interdisciplinary integration within its Office of Research and Development. This article details the history of this process and discusses lessons learned from other federal agencies seeking to integrate social and biophysical research: finding the right combination of top down and bottom up approaches; balancing objectives of advancing science and/or supporting programmatic operations; using social science methods to inform the process; and engaging multiple stakeholders. Attention to the social context of scientific practice, including research processes and research use, fosters success. Three strategies for integrating social sciences to support solutions-driven environmental research are: weaving social science throughout the research process, strengthening social networks, and fostering interdisciplinary hubs. Integration into planning and carrying out research has greater transformative potential than integration into product development and distribution. This article provides insights into institutional considerations for advancing interdisciplinarity and the social context of scientific practice in government agencies. It illustrates the multiple decision contexts and inclusion of social science at the science-policy interface.

1. Introduction

The U.S. Environmental Protection Agency (EPA) works to solve complex environmental problems to protect human health and the environment. These problems may involve multiple contaminants, media, and ecosystems; travel circuitous exposure pathways that involve human behavior and built environments; and reflect historical legacies as well as contemporary environmental change. The problems are entangled with social, political, and economic dynamics in diverse cultural contexts. Solving complex environmental problems requires research that integrates the social and biophysical sciences (Brondizio et al., 2016; Burke et al., 2017; Clark and Wallace, 2015; National Research Council, 2012). Risk assessments, for example, while grounded in toxicology, also benefit from the social sciences. Assessing children’s exposure to lead or effects of air pollution on cardiovascular health necessitates behavioral, epidemiological, public health, sociological, and engineering research to understand who is most at risk, how, and why (Burke et al., 2017; Gwinn et al., 2017; Zartarian et al., 2017). Studies of environmental justice, interactions between chemical and non-chemical stressors, and social determinants of health (e.g., housing quality, health care access) support risk assessment and management (Birnbaum et al., 2016; Braveman and Gottlieb, 2014).

This article chronicles how the EPA’s Office of Research and Development (ORD) is working to integrate the social and biophysical sciences to advance interdisciplinary, solutions-driven science for environmental problem solving. In this paper, the term “social science research” encompasses the use of theory and methods from diverse fields (e.g., anthropology, economics, geography, political science, sociology). The article draws upon work on interdisciplinarity, the social production of science, and the science-policy interface to advance knowledge about institutional considerations in interdisciplinary integration.

First, institutional background on ORD is provided along with a brief history of social science in the organization. Next, key takeaways from other federal agencies are reviewed: balancing top down and bottom up approaches; using social science research to advance general scientific knowledge and/or support an agency’s programmatic operations; using social science methods to inform integration; and engaging multiple stakeholders. Finally, details are provided on the initiative’s three strategies for success: integrating social science throughout the research process, strengthening social networks, and fostering interdisciplinary hubs. The research process has four stages: planning and carrying out research, and designing and distributing products. At each stage, the social context of scientific practice is considered. Integration into planning and carrying out research is likelier to induce transformative change; integration into product development and distribution offers primarily additive benefits. The article provides insights into institutional considerations for interdisciplinarity and the social context of the production of science in a federal agency. Lessons learned from this initiative may be of use to other organizations seeking to follow a similar path.

The connection between the theories and practices of interdisciplinarity needs strengthening (Banner, 2013). On the theoretical side are proposed conceptual approaches and analysis of epistemological barriers (Banner, 2013; Brondizio et al., 2016; MacMynowski, 2007; Palsson et al., 2013; Wagner et al., 2011. On the practical level are suggestions for team and researcher characteristics and processes (Brown et al., 2015; European Science Foundation, 2013; Fischer et al., 2011; Palmer et al., 2016). This article builds on existing research on the institutional context of interdisciplinary integration (Fischer et al., 2011; Gardner, 2013). It blends the theoretical and the practical in its analysis of the social context of operationalizing interdisciplinarity in one organization, ORD.

This article also contributes to analysis of the science-policy interface. Recent literature has focused on the importance of building networks and relationships and the role of organizations in spanning the science-policy boundary (Dunn et al., 2017; Posner and Cvitanovic, 2019; Sarkki et al., 2019). It emphasizes the importance of including social science and transdisciplinary research at this interface (Choquet et al., 2018; Sarkki et al., 2019; Wesselink et al., 2013), but doesn’t necessarily detail how to do so or what it means. This article provides concrete examples of linking science and policy through interdisciplinary, solutions-driven research. In addition, it expands the notion of policy beyond formal regulations to multiple decision types. As such, it helps flesh out the ‘complex ecosystem’ of boundary spanning science-policy (Posner and Cvitanovic, 2019).

2. Methods

Research for this article was based on qualitative content analysis of materials related to social science integration produced about, by, or for federal agencies (Babbie, 1998; Merriam and Tisdell, 2016; Spencer, 1994). Qualitative content analysis entails critical attention to language and identification of patterns. We used publicly available sources: reports, memos, communication materials, published research, and planning documents. Initial sources were found using recommendations and thematic website searches, snowballing from there. Triangulation of materials from multiple sources and peer examination by ORD staff contributed to the study’s validity and reliability (Babbie, 1998; Merriam and Tisdell, 2016).

Another important aspect of methodology is the researchers’ standpoint, or perspectives, goals, and orientation towards the problem at hand (Clark et al., 2017). Our standpoint as authors who work in ORD is that of an institutional insider (Spencer, 1994; Wolcott, 1995). This study is informed by our in-depth understanding of how the organization works on a daily basis. Our personal connections in EPA and other agencies helped us identify sources and compare situations. Most importantly, we are not merely observing social science integration, we are also working to enact it.

3. Social and environmental sciences in ORD

As the science arm of the agency, ORD works closely with EPA program offices that regulate air, water, waste, and chemical safety, and with EPA regional offices that work with state and local agencies to implement environmental laws and policies. ORD research informs regulatory decisions, contaminated site cleanups, water treatment technologies, and other aspects of the agency’s mission. ORD scientists work in laboratories and research centers across the country. They are trained in disciplines as diverse as computational toxicology, chemistry, microbiology, ecology, economics, and engineering. ORD research is coordinated through six National Research Programs: air and energy, chemical safety for sustainability, homeland security, human health risk assessment, safe and sustainable water resources, and sustainable and healthy communities. The National Research Programs are by nature interdisciplinary, although the social sciences have played a limited role to date. There has been growing recognition in the agency of the need for solutions that address the complexity of environmental problems and decision contexts. Solutions-driven research integrating social and environmental science is one way ORD works to meet this need.

EPA’s program and regional offices engage in social science through regulatory cost-benefit analysis and policy evaluation. In 1971, EPA established an environmental economics program. Economic analysis became more central to the agency during the Reagan and George H.W. Bush administrations after Executive Order 12291 and increasing emphasis on market incentives (McGartland, 2013). Its Office of Policy undertook more economics work and established the National Center for Environmental Economics (McGartland, 2013). Incorporation of other social science disciplines happened more slowly. In 1996, the Office of Sustainable Ecosystems and Communities entered into a cooperative agreement with the Society for Applied Anthropology to analyze human dimensions of environmental problems (Society for Applied Anthropology, 1996 [online]). In 2015, Executive Order 13707 encouraged federal agencies to use behavioral and social sciences in decision-making (3 CFR 56365-56367, 2015). The order spurred activities in several agencies, including EPA.

The external and internal impetus for integrating social and environmental science in ORD can be traced back to the environmental laws and policies underlying its research mandate (Table 1). Support for environmental social science begins with the National Environmental Policy Act of 1969, which states that federal agencies shall “utilize a systematic, interdisciplinary approach which will insure the integrated use of the natural and social sciences and the environmental design arts in planning and in decision-making” (42 USC1970 § 102(2)A). Executive branch policies also have shaped ORD’s research directions. For example, the September 11, 2001, terrorist attacks and the Amerithrax incident sparked the creation of ORD’s National Homeland Security Research Center. At first a short-term enterprise, the center was made permanent as White House policies expanded EPA’s roles in disaster preparedness and response (Presidential Policy Directive-8, 2011; Presidential Policy Directive-21, 2013). Its research now addresses social and technical aspects of resilience.

Table 1.

Key Influences and Efforts to Integrate Social Science in ORD.

1969	National Environmental Policy Act
1971	ORD creates Environmental Economics Research Program
1981	Executive Order 12291
1995	STAR grants for social science
1996	Cooperative agreement with the Society for Applied Anthropology
2005	Environmental economics research strategy
2011	ORD town hall on social science research needs
2011	Scientific Advisory Board and Board of Scientific Counselors OSC recommendations on EPA social science
2012
2015
2017
2012	National Research Council recommendations on EPA social science
2012	ORD creates Sustainable and Healthy Communities research program
2015	Executive Order 13707
2016	ORD holds first social science “bootcamp” for staff; Social Environmental Science Exchange community of practice established
2017	ORD hires a full-time social science advisor

Two independent federal advisory panels recommended enhancing social science research in ORD. The Scientific Advisory Board (SAB) provides scientific and technical advice to the EPA administrator. The Board of Scientific Counselors (BOSC) gives recommendations to ORD. For some years, the SAB and BOSC have urged ORD to use more social science to formulate problems, conduct systems research, and develop solutions that account for human behavior (Scientific Advisory Board, 2010; Scientific Advisory Board and Board of Scientific Counselors, 2011). In 2011, the panels recommended that ORD “take specific steps to enhance its expertise and research” in social sciences (Scientific Advisory Board and Board of Scientific Counselors, 2011). They underscored this point with more detailed suggestions in 2012 and 2015 (Scientific Advisory Board and Board of Scientific Counselors, 2012, 2015), and reaffirmed this need in 2017 (Board of Scientific Counselors, 2017). The National Research Council and National Academies of Sciences have endorsed social science’s role in environmental research generally (National Academy of Sciences, 2017b; National Research Council, 2005), and EPA specifically (National Research Council, 2012).

ORD has undertaken isolated efforts to include social science. Between 1995 and 2015, Science to Achieve Results (STAR) research grants and graduate fellowships funded social science projects, although science and engineering grants predominated (National Academy of Sciences, 2017a). ORD and the Office of Policy co-produced a research strategy for environmental economics in 2005 (EPA, 2005). In 2011, ORD assessed social science research needs through a town hall, community meetings, and interviews. Participants agreed on the need for behavioral and social science research and discussed areas of interest (Daniels et al., 2011). In 2012, ORD created the Sustainable and Healthy Communities Research Program and conducted a workshop to identify core social science competencies. As a result, new research areas (e.g., ecosystem goods and services, sustainability) were created.

These efforts waxed and waned as research needs, key personnel, budgets, and resources changed. A more comprehensive approach has slowly emerged. National Research Programs emphasized the role of social science in understanding environmental problems, analyzing decision-making, and applying knowledge to benefit society (Costa and Hubbell, 2016; EPA, 2016a, b). ORD hired a full-time social science advisor in 2017, enabling an organization-wide perspective. The concurrence of external mandates, leadership support, and staffing capacity spurred the current initiative to strengthen social-environmental science in ORD’s research portfolio.

4. Social context of ORD science

Social sciences enrich environmental research by analyzing social phenomena that cause environmental change or mediate its impacts, examining people’s lived experiences with environmental problems, providing insights into human behavior, and bringing societal outcomes of environmental policy to light (Barnes et al., 2013; Clayton et al., 2016; Hackmann et al., 2014; Maxwell, 2014; Weaver et al., 2014). Table 2Table Two details potential contributions of social science to solutions-driven research in ORD.

Table 2.

How social science supports solutions-driven research.

It improves	Because it helps
Case study research	Produce generalizable knowledge from case studies.
Environmental justice	Analyze socioeconomic and political dimensions of risk distribution.
	Provide a framework for community engagement.
Environmental modeling	Incorporate social dynamics and human behavior to improve characterization of ecological and health endpoints.
Human health and exposure research	Specify how social and environmental stressors interact to affect chemical exposure and human health.
Participatory research	Involve stakeholders to build trust, increase scientific literacy, and produce usable results.
Research planning	Clarify the problem to study and provide a more comprehensive framing.
Risk communication	Develop and test messaging strategies.
	Consider the social context of communication.
Solutions	Identify actions EPA programs and regions could take & evaluate effectiveness.
Systems research	Analyze social-environmental systems.
Tool development & use	Design tools that align with decision-making.
	Develop place-based applications for tools.

Science and technology studies literature shows that organizational structure, cultural values, cognitive models, epistemological frameworks, and social interactions affect research decisions and outcomes (Jasanoff, 1992; Fortun et al., 2016; Latour, 1988; Lélé and Norgaard, 2005; Taylor, 1992; Wynne, 1996). Taking this social context of scientific practice into consideration is central to the success of this initiative. To do so, the initiative approaches ORD research as a process divided into four stages: planning, carrying out research, product development, and product distribution. The stages may be sequential or overlap. This internal process intersects with research use, or how society at large generates the need for, understands, uses, and benefits from research.

5. Lessons learned from other U.S. federal agencies

This initiative benefits from efforts in other U.S. federal agencies to integrate social science (Table 3). An attempt to trace the arc of these efforts and their cross-pollinations is outside the scope of this paper as they have varied influences and developmental paths. Four key takeaways from these efforts are: finding the right combination of top down and bottom up institutional approaches, ascertaining whether the objective is to advance general scientific knowledge and/or support programmatic operations, using social science methods to inform the integration process, and engaging multiple stakeholders.

Table 3.

Select U.S. federal agency social science programs in 2018.

Federal Agency	Social Science Program
U.S. Department of Agriculture
Forest Service	Research Stations- units and individuals
National Institute of Food and Agriculture	Throughout
Economic Research Service	Throughout
U.S. Department of Commerce
National Oceanic and Atmospheric Administration	Performance, Risk and Social Science Office
National Ocean Service	Center for Ocean Science- teams and individuals
U.S. Department of Health and Human Services
National Institutes of Health	Office of Behavioral and Social Sciences Research
National Institute of Environmental Health Sciences	Partnerships for Environmental Health
Centers for Disease Control and Prevention	Behavioral and Social Sciences Working Group
U.S. Department of the Interior
Bureau of Land Management	Socioeconomics Program
National Park Service	Social Science Branch
United States Geological Service	Social and Economic Analysis Branch
General Services Administration	Office of Evaluation Sciences
U.S. Global Change Research Program	Social Science Coordinating Committee

Some agencies take a top down approach where integrating social science is guided by a strategic vision or plan. This is the case in the National Institutes of Health’s Office of Behavioral and Social Sciences (Office of Behavioral and Social Sciences, 2016), National Oceanic and Atmospheric Administration (National Oceanic and Atmospheric Administration Social Science Committee, 2015), U.S. Forest Service (U.S. Forest Service, 2004), U.S. Global Change Research Program (U.S. Global Change Research Program, 2012), and Bureau of Land Management (Bureau of Land Management, 2013). A top-down approach can jumpstart agency-wide change and be used to track progress. Creating a far-reaching strategy takes time, resources, and institutional commitment, though. It runs the risk of being too high level to affect programmatic change and is subject to shifts in organizational priorities Table 3.

In a bottom-up approach, offices or individuals from across the agency take the lead. For example, some of the U.S. Forest Service’s seven research stations have designated social science units; others house social scientists in multidisciplinary units. A bottom-up approach can shorten the time frame to producing outputs, build relationships among staff, and allow offices to proceed at their own rate. It runs the risk of being too piecemeal and creating redundancies. ORD balances top-down and bottom-up approaches by fostering interdisciplinary hubs.

Another difference is the degree to which social science research is used to advance general scientific knowledge and/or support programmatic operations. U.S. Forest Service research stations and the U.S. Geological Survey’s Social and Economic Analysis Branch, for example, investigate urban forestry, ecosystem services, and human dimensions of resource management. These examples fit under the umbrella of advancing scientific understanding through basic or applied research. In contrast, the National Park Service’s Social Science Branch collects visitor statistics and evaluates customer service to aid Park Service planning. Similarly, the National Oceanic and Atmospheric Administration’s Performance, Risk and Social Science Office assesses the agency’s impact on society. Both use social science to support programmatic objectives. The main objective of the ORD initiative is to conduct research to advance scientific understanding and solve environmental problems. Achieving this goal may require programmatic applications of social science to inform how the organization carries out this work.

A second lesson learned is to utilize social science research methods as part of the integration process. The Bureau of Land Management assessed its social science needs and capabilities through interviews and surveys (Bureau of Land Management, 2013). The National Institutes of Health did concept mapping and held interviews to produce its prospectus for social science research (Office of Behavioral and Social Sciences, 2007). Three Department of Interior agencies surveyed federal and state agencies on social science capacity and resource needs (Sexton et al., 2013). In a similar vein, ORD and the National Center for Environmental Economics interviewed staff to develop an Environmental Economics Research Strategy (EPA, 2005). More recently, ORD mapped its social networks and surveyed staff to identify areas of interest and expertise in social science. It used participatory observation to understand how social interactions among scientists affect interdisciplinary problem formulation (EPA, 2017).

A final lesson is the importance of engaging multiple stakeholders (Fiske, 2014). The National Institutes of Health held a town hall meeting to solicit feedback on its draft prospectus (Office of Behavioral and Social Sciences, 2007). The National Oceanic and Atmospheric Administration’s Research Council established a Social Science Committee with representatives from different offices (National Oceanic and Atmospheric Administration Science Advisory Board, 2009). ORD held a social science boot camp for scientists in 2016 and established an agency-wide Social Environmental Science Exchange network.

6. Three key elements of the ORD approach

6.1. Weave it through the research process

The first key element of ORD’s approach is to weave social science throughout the four stages of the research process (Fig. 1). Doing so makes opportunities for improvement clear and distinguishes transformative from additive benefits. The initiative considers the social context of scientific practice in each stage. Social science does not need to be included in all stages of every project. For some projects, product development may be more relevant; for others, it is research planning.

Fig. 1.

Research process stages. The four research stages in ORD offer distinct opportunities and strategies for engaging social science. The effects of doing so range from additive to transformative.

6.1.1. Plan research

Integrating social science into research planning, as recommended by EPA advisory panels (Scientific Advisory Board and Board of Scientific Counselors, 2011), requires changes to scientific practice. Codifying social science, that is, making it part of formal institutional documents and procedures, provides the greatest opportunity for long-term success (Fiske, 2014). In ORD, codification occurs through quinquennial Strategic Research Action Plans. In the most recent round of planning, ORD’s social science advisor and social scientists on staff reviewed the plans and recommended language changes. The plans now reflect social science research questions and interdisciplinary approaches.

ORD also encourages new research directions through innovation funding. It is a less sure route to codification but can be an effective way to test new ideas. ORD’s 2017 through 2019 calls for State and Regional Innovation Proposals identified social sciences as a priority area and funded three projects: interviewing farmers about nutrient management decisions (EPA, 2018a), communicating indoor air quality risk reduction practices following flooding (EPA, 2019a), and evaluating messages to decrease air pollution from wood stoves (EPA, 2019b).

Social science integration into research planning benefits from documenting recent advances and gaps to identify topics to target (EPA, 2016b). Techniques for doing so include evidence gap maps, systematic reviews of existing studies, meta-analysis, and meta-ethnography (McKinnon et al., 2015; Noblit and Hare, 1988; Singh, 2017). The Homeland Security Research Program, for example, did a systematic literature review of the social science of environmental cleanups (Maxwell et al., 2018). The Air and Energy Research Program identified knowledge gaps about social factors influencing air quality sensor use (Hubbell et al., 2018).

Integrating social and environmental science in research planning for individual projects requires attention to problem formulation, or clarifying the problem to study (Fig. 1) (Clark et al., 2017; EPA, 2016b). Contextualizing environmental problems within societal dynamics and stakeholder perspectives can re-orient or expand initial research. For an EPA-sponsored workshop on wildfire smoke health risk communication, for example, participants generated mind maps, or diagrams of how topics are interrelated, which formed the basis for small group discussions to refine problem statements (EPA, 2017).

The social context of research planning can present challenges to social science inclusion. Competition for scarce budgets or conflicts over project ownership can surface. Differences in scientific paradigms and epistemology, or the assumptions disciplines have about how to produce knowledge and understand the world, also arise (Clark and Wallace, 2015; European Science Foundation, 2013; Moon and Blackman, 2014). Disciplinary hierarchies and power differentials come into play (Gardner, 2013; Ledford, 2015; MacMynowski, 2007).

There are ways to resolve differences and build trust (Clark et al., 2017; Eigenbrode et al., 2007; Harris and Lyon, 2013; Olabisi et al., 2014). Interdisciplinary concept mapping and interdisciplinary context mapping provide bridges across disciplines (Clark et al., 2017; Heemskerk et al., 2003; Morse, 2014). For example, ORD used the DPSIR (Driving Forces – Pressures – State – Impacts – Responses) framework to develop a research plan on water quality issues in Puerto Rico (EPA, 2015b).

6.1.2. Carry out research

Carrying out interdisciplinary research may involve asking new research questions within an integrated problem frame or adding socioeconomic elements to a technical area of study. One ORD project investigates social acceptance of technology alternatives for addressing nutrient pollution in Cape Cod, Massachusetts (Chintala et al., 2018). In the Great Lakes, researchers did participant observation on multi-stakeholder decision making for environmental remediation (Williams et al., 2018).

Qualitative and quantitative methods, datasets, and scales of analysis used in the social sciences complement natural and physical science techniques. Case study research generates an in-depth understanding of environmental problems within a social context. Recent advances in multi-method research and data synthesis techniques (e.g., agent-based modeling, scenarios, games) aid analysis of complex society-environment dynamics (Filatova et al., 2013; Palmer et al., 2016; Rounsevell et al., 2014). However, the time required for social data collection and analysis, particularly when using ethnographic methods, can be a barrier to adding social science to time-sensitive research (e.g., decontamination after a chemical spill, addressing emerging contaminants).

Attention to the social context of carrying out research can smooth the path forward. Necessary authorization (e.g., human subject research approvals) must be obtained before fieldwork begins. Additionally, federal agencies require permission from the White House Executive Office of Management and Budget to collect data from members of the public. Interdisciplinary research teams can take measures to improve internal communication, collaborative workspaces, social interactions, and decision processes (EPA, 2016b; Palmer et al., 2016; Read et al., 2016; Stokols et al., 2008). Another aspect of ORD’s social context is that producing high quality science is a bedrock of its institutional identity. Its stringent quality assurance procedures help sustain this identity (c.f. Jasanoff, 2011). Quality assurance concepts (e.g. rigor, uncertainty) must be translated into quality control techniques that make sense for qualitative methods (Lamont and White, 2005; Social Science Research Council, 2018).

6.1.3. Develop and distribute products

ORD products are intended for scientific audiences (e.g., journal articles, reports, models, datasets), applied decision-making (e.g., tools, technical briefs), and direct applications (e.g., monitoring or environmental compliance technologies). To date, ORD has developed several products with social science content. The Decision Analysis for a Sustainable Environment, Economy and Society (DASEES) tool supports multi-stakeholder, multi-criteria decision processes for managing natural resources (Stockton et al., 2012). The triple value model addresses social, economic, and environmental dimensions of sustainability (Fiksel et al., 2014).

Social science concepts and methods can inform programmatic operations of product development. Interdisciplinary interactions aid design of products applicable to a variety of stakeholders and decision contexts (Hoover et al., 2015). Human-centered design borrows methods from the social sciences to identify end-user needs when designing a product, and later in testing it based on end-user experiences (Baxter et al., 2015). In addition, when product development is grounded in theories of behavioral change, it may increase the impact of products in achieving institutional or individual change that can improve environmental quality and public health.

A social context challenge of this stage stems from graduate training in the social sciences that focuses on producing journal articles and books, not decision-support tools. Similarly, while some disciplines such as environmental economics are accustomed to modeling, others such as anthropology and sociology are not. Indeed, sociologists and anthropologists may experience discomfort translating nuanced perspectives on communities, culture, and power into a discrete set of variables to input into a model or tool.

EPA has made progress in linking social and natural science models. In its recent review of the national ambient air quality standards for ozone, Community Multiscale Air Quality (CMAQ) model outputs were used in combination with the Forest and Agricultural Systems Optimization Model (FASOM) to estimate economic impacts of meeting alternative ozone standards (EPA, 2015c). Advances in data integration have been slower. One way forward might be for ORD to integrate multidisciplinary datasets into EnviroAtlas (Pickard et al., 2015) and its other Geographic Information Systems (GIS) resources. Joint longitudinal databases linking health and environmental data with behavioral and socioeconomic data could also enrich interdisciplinary research infrastructure.

Product distribution benefits from social science concepts and methods to communicate results, connect science with decision-making, and evaluate product utility. Risk communication and risk perception science provide strategies for testing how to best communicate risk and uncertainty in meaningful ways with different audiences (Glik, 2007; Horlick-Jones et al., 2003; Slovic, 1987). Interdisciplinary research teams can translate research to create solutions and mitigate risks (Anderson et al., 2015). The social context of product distribution is complicated, though, with multiple EPA offices involved. ORD communications staff lead social media efforts. Program and regional offices serve as intermediaries with stakeholders. ORD scientists may not be embedded in social networks alongside prospective audiences and end users.

Social science can enhance ORD products’ ability to inform decisions and enact solutions that meet feasible, appropriate, meaningful, and effective (FAME) criteria for evidence-based practices (Jordan et al., 2019). By informing FAME solutions, integrated social and natural science research support stakeholder acceptance of decisions and achieve lasting improvements in environmental quality and public health. Recent examples include working with urban planners to use Health Impact Assessments to increase net benefits of development projects (EPA, 2015a); and developing training and outreach to cardiologists and cardiovascular patients to increase awareness of risks from air pollution (Baghdikian et al., 2019).

6.1.4. Transformative and additive benefits of social science in the four stages

Integrating social science into planning and carrying out research has the potential to transform outcomes for a given project and the research process itself (Strang and McLeish, 2015). Applying social science to product development and distribution adds value to specific outputs but is less likely to stimulate far-reaching changes to scientific practice. Its benefit is thus additive rather than transformative. There are exceptions to this pattern. Employing human-centered design can potentially transform product development by considering research use from the onset.

Distinguishing transformative from additive benefits can help non-social scientists realize that purely additive measures should not be conflated with complete integration of social science. Using communication science concepts to better communicate risk levels established through toxicological research is of value. It is not the same, however, as investigating how stakeholders perceive risks and how environmental risks are embedded in socioeconomic contexts.

The transformative potential of social science in the latter two stages can be enhanced through institutional shifts toward solutions-driven research, coproduction of knowledge, and stakeholder engagement. Translating research results into actionable information is most effective when it begins with problem formulation (Burke et al., 2017). Social-environmental science integration thus is an iterative process that takes place across several research cycles.

6.2. Strengthening social networks

The second key element of the ORD initiative is to develop a network of EPA staff trained or interested in social science (EPA, 2016b). It established the Social Environmental Science Exchange to build capacity in social science and foster relationships among ORD, program office, and regional staff. Since its inception in August 2016, over 270 EPA staff members have joined the network. The Social Environmental Science Exchange holds monthly teleconference calls and webinars that usually reach 70 to 90 participants from around the agency. Discussion topics to date include community-based social marketing for environmental behaviors, ethnographic analysis of energy use, social determinants of health, environmental justice, and application of social science to regulatory development and enforcement. Presenters come from EPA, other government agencies, and academia. Network organizers manage an internal, online toolbox with materials on social science frameworks and methods, links to key databases, software tools, models, and training. The toolbox has a discussion forum for member-initiated, informal conversations. A next step in strengthening social networks could be to expand the Social Environmental Science Exchange to include non-EPA members, gaining additional input into problem identification and product design.

This network generates the connective tissue to sustain interdisciplinary hubs. One way it has done so is through three pilot Social-Environmental Science (SES) Dialogues to crowdsource advice for ORD research projects seeking to integrate social science. The SES Dialogues are modeled on the Resilience Dialogues, a public-private partnership that engages community leaders and scientists on climate change adaptation (Resilience Dialogues, 2017). In the SES Dialogues, ORD project teams and volunteer subject matter experts from in and outside the agency participated in a facilitated, two-week long, online, asynchronous conversation. The dialogues clarified the problem to be studied and identified relevant theories, methods, and research questions. One dialogue discussed methodologies for eliciting stakeholder values about changing waterbody conditions related to nutrient levels. Another considered ways of analyzing the health effects of chemical and non-chemical stressors. The next round of dialogues will address how social science can inform solutions to pressing problems in EPA regions.

6.3. Foster interdisciplinary hubs

The third key element is to balance top down and bottom up approaches by fostering interdisciplinary hubs. Not all research projects are well-suited for including social science. Barriers include: the topic is too far removed from its societal context (e.g., molecular toxicology research on chemical effects at the cellular level); scientists are not eager to expand a project’s focus or do not know where to start; the need to share an ever-shrinking pie of resources; and lack of access to necessary disciplinary expertise. ORD addresses these challenges by using interdisciplinary hubs to build capacity at the research project level and encourage top-down leadership support and institutional codification. Fostering interdisciplinary hubs also helps overcome staffing limitations. Including social scientists on research teams helps ensure that social science is incorporated into all four stages of the research process. ORD’s staffing capacity is limited, though. As of 2018, around 3.4 percent of ORD scientists had graduate or undergraduate majors related to social science. However, fewer than two percent of ORD federal staff held e job series in social science (e.g. Economist, Anthropologist) signifying that they are expected to do this kind of work. ORD augments limited staff time through term-limited appointments (e.g., postdoctoral fellows).

Interdisciplinary hubs are beginning to converge around nutrients in aquatic ecosystems, interaction of social and environmental stressors, wildfire smoke and human health, resilience, and environmental cleanups. They provide a decentralized forum for experimentation that has access to centralized support and the collective insights of the Social Environmental Science Exchange. Interdisciplinary hubs can benefit from conceptual frameworks for social-environmental systems analysis (e.g. Fiksel et al., 2014; Gee and Payne-Sturges, 2004; Tulve et al., 2016). Reconfiguring research infrastructure (e.g., hiring, contracts, funding, software, professional advancement) sustains long term support for interdisciplinary integration (EPA, 2016b; Maxwell, 2014). Contract vehicles for extramural research funding might not all have social science expertise, for example. Providing computational support along with software aids integration of social and environmental data (Palmer et al., 2016). Also important is training staff in new skills such as team science (Ledford, 2015; National Research Council, 2015; Read et al., 2016).

ORD is a large and widespread organization, where members of the same research team may work in different states. A key challenge is how to replicate successful integration strategies (and avoid pitfalls) from one hub to another. Existing organizational resources may be of help. The Regional Sustainability and Environmental Sciences Research Program applies ORD science to environmental problems in EPA regions, resulting in research that is transferable from one location to another. In one case, ORD researchers, the Proctor Creek, Atlanta community, and EPA Region 4 developed methods for conducting health impact assessments (EPA, 2015a). Other EPA programs and regions are now replicating these methods. The Social Environmental Science Exchange also offers venues for sharing successful strategies.

7. Conclusions

Integrating biophysical and social science research to solve environmental problems is a long-term endeavor. ORD has benefitted from lessons learned at other federal agencies seeking to do the same. Lessons center on finding the right combination of top-down and bottom-up institutional approaches, determining if the objective is to advance scientific understanding and/or provide programmatic support, using social science methods to inform the integration process, and engaging multiple stakeholders.

ORD merges top-down and bottom-up approaches by undertaking organization-wide and project-level actions. Three key elements of its approach are to employ social science in all four stages of the research process, strengthen social networks, and foster interdisciplinary hubs. Attention to the social context of scientific practice is critical to success. Integrating social science into planning and carrying out research offers the greatest potential for transformative change. Using social science insights to develop and distribute products adds value to research outputs and may be appropriate for some projects. It is not a substitute for substantive modifications to research design and methods.

Merging top-down and bottom-up approaches by fostering interdisciplinary hubs and strengthening social networks allows for institutional codification, reconfiguration of research infrastructure, and organizational cultural change. Norms of scientific practice, disciplinary hierarchies, organizational identity, resource allocation, and social relationships do not change overnight. Changing research processes takes substantial time and energy; addressing research use is more complicated still. Initial signs of progress are recent social science publications (e.g. Merrill et al., 2018; Maxwell et al., 2018), sustained interest in Social Environmental Science Exchange webinars, and participation in the SES Dialogues.

This article presents strategies for institutional integration of social and biophysical science research for interdisciplinary environmental problem solving. It provides insights into the social context of scientific practice. The integration process is not a linear one. It exhibits a certain inherent messiness and is replete with ebbs and flows. Limitations to replicability and scalability persist, including staffing capacity and finite resources. Integrating social and environmental science is an iterative process that takes place across several research cycles but has the potential to be of long-term benefit to protecting human health and the environment.

Integrating social sciences into EPA’s scientific research portfolio can increase scalability of research results for use in policy and decision-making at multiple levels (e.g., individual, community, state, federal). Social science methods can be used to address the context of decision-making at these various levels. For example, insights from behavioral science and theories of change can help design research to inform individual decisions to reduce exposure and improve health. Place-based insights from ethnographic studies can help improve community uptake of science-based solutions by accounting for social and cultural factors.

ORD’s experiences have implications for analyzing the science-policy interface writ large. This article expands the notion of “policy” to include diverse decision contexts. ORD’s science can inform EPA regulatory decisions. It also has importance for emergency responders, state environmental agencies, watershed management associations, physicians, air quality boards, and many other social actors. This article also illustrates what can happen when adding social sciences to the science-policy interface. Social scientists provide methods and insights that improve understanding of decision contexts at multiple scales and forms. Better understanding of the context of research use can help ORD assess research needs, design projects, engage stakeholders, provide technical assistance, and deliver products in appropriate formats—all elements of solutions-driven research.

Highlights.

• Solving environmental problems requires social and environmental science research.

• Social contexts of scientific practice shape how agencies carry out social science.

• Attention to institutional research processes aids interdisciplinary integration.