Setting up a Horizon Scanning System

Home , Horizon scanning, Technological change, Technology assessment

WFRXXX10.1177/1946756717749613World Futures ReviewHines et al. 749613research-article2018

Article

World Futures Review 2018, Vol. 10(2) 136–151 Setting Up a Horizon Scanning © The Author(s) 2018 Reprints and permissions: sagepub.com/journalsPermissions.nav System: A U.S. Federal Agency https://doi.org/10.1177/1946756717749613DOI: 10.1177/1946756717749613 Example journals.sagepub.com/home/wfr

Andy Hines1, David N. Bengston2, Michael J. Dockry2, and Adam Cowart1

Abstract Managers and policy makers are continually working toward a desired future within a context of rapid and turbulent change. To be effective in this context, they must look ahead to anticipate emerging trends, issues, opportunities, and threats. Horizon scanning is a foresight method that can help managers and policy makers develop and maintain a broad and externally focused forward view to anticipate and align decisions with both emerging (near-term) and long-term futures. This article reports on the design and early stage development and implementation of a horizon scanning system established for the U.S. Forest Service (USFS), Strategic Foresight Group, and developed cooperatively with the University of Houston Foresight Program. The goal of the project is to develop an ongoing horizon scanning system as an input to developing environmental foresight: insight into future environmental challenges and opportunities, and the ability to apply that insight to prepare for a sustainable future. In addition, the horizon scanning system is supported by volunteers from within the Forest Service. By including participants from throughout the Forest Service, the project seeks to foster a culture of foresight within the organization, and eventually to develop a more forward looking organizational structure for the USFS and other natural resource management agencies. Lessons learned from the experience to date are shared as well as future challenges for keeping the horizon scanning system in good working order—current, relevant, and consistent.

Keywords foresight, horizon scanning, forestry, emerging issues, policy

Introduction as “the acquisition and use of information about events, trends and relationships in an organiza- Managers and policy makers are continually tion’s external environment, the knowledge of working toward a desired future within a con- which would assist management in planning the text of rapid and turbulent change. To be effec- organization’s future course of action” (Choo tive in this context, they must look ahead to 2002, 84). anticipate emerging trends, issues, opportunities, and threats. Horizon scanning (also referred to as environmental scanning) is a foresight 1University of Houston, Houston, TX, USA method that can help managers and policy mak- 2U.S. Forest Service, St. Paul, MN, USA ers develop and maintain a broad and externally Corresponding Author: focused forward view to anticipate and align Andy Hines, University of Houston, 4235 Cullen Blvd., decisions with both emerging (near-term) and Cameron Building, Houston, TX 77204, USA. long-term futures. Horizon scanning is defined Email: [email protected] Hines et al. 137

Horizon scanning has long been practiced inaccurate expectations of shoreline erosion in the military, the intelligence community, rates, and overoptimistic cost estimates of arti- and the business world. Business research has ficial beaches (Pilkey and Pilkey-Jarvis 2007; demonstrated that effective horizon scanning Sarewitz et al., 2000). improves organizational performance (Choo The inability to accurately forecast the 2002). In recent years it has been increasingly future of complex social-ecological systems used in other public sector fields, such as stems from many factors, such as the emergent human health (Douw and Vondeling 2006) and nature of these systems and their sensitive education (Munck and McConnell 2009). But dependence on initial conditions. Emergent the use of formal horizon scanning in natural properties of a system cannot be predicted from resources and environmental organizations has the parts (de Haan 2006). The prevalence of sur- been limited (Sutherland and Woodroof 2009). prise in social-ecological systems (Gunderson Among the few examples are scanning exer- and Longstaff 2010) implies that some impor- cises related to biodiversity (Sutherland et al. tant uncertainties in long-term forecasts are 2008), global conservation issues (Sutherland irreducible and that traditional scientific tools et al. 2015), and environmental issues with and models are blunt instruments for studying implications for the U.S. military (U.S. Army a future that does not exist. Environmental Policy Institute n.d.). The goal of horizon scanning, therefore, is The Millennium Ecosystem Assessment not to try to accurately predict the future, but to concluded that “over the past 50 years, identify, compile, and analyze the various sig- humans have changed ecosystems more rap- nals of change. These signals can be explored idly and extensively than in any comparable to stimulate thinking about the range of future period of time in human history, largely to possibilities. While getting “the future” pre- meet rapidly growing demands for food, fresh cisely right is futile, it is possible to use signals water, timber, fiber, and fuel” (Millennium of change identified in horizon scanning to Ecosystem Assessment 2005, p. 1). Surprises map plausible scenarios and pathways to the are increasing along with the expanding scale future (Hines 2009). Horizon scanning serves of human impacts (Gunderson and Longstaff as an early warning system to identify poten- 2010). Hibbard et al. (2007) refer to the period tial threats and opportunities. Some character- following World War II as the “Great istics of horizon scanning that distinguish it Acceleration,” a time of significant increase in from typical activities that planners use to sur- the scope, scale, and intensity of impacts on vey future conditions include (Bengston 2013): the social-ecological system. At the same time, the rapid pace of technological change has the •• It emphasizes “weak signals” or early potential for sweeping environmental effects, indicators of potential change, such as genetic engineering and nanotechnolo- •• It is comprehensive across a wide range gies, and makes the future landscape more of sectors, challenging to discern (Olson and Rejeski •• It emphasizes external trends and devel- 2005). opments, and In the face of such profound and uncertain •• It includes possible wild cards or low- change, foresight developed through horizon probability, high-impact events. scanning is not an attempt to “predict the future.” It is simply not possible to accurately This article reports on the design and early predict the future of complex social-ecological stage development and implementation of a systems for long timeframes. Efforts devoted horizon scanning system established for the to social and environmental prediction have U.S. Forest Service (USFS), Strategic been plagued by a host of shortcomings, and Foresight Group, and developed cooperatively unjustifiable faith in these predictions have with the University of Houston (UH) Foresight contributed to unmanageable nuclear waste Program. The goal of the project is to develop disposal practices, poisoned mining sites, an ongoing horizon scanning system as an 138 World Futures Review 10(2) input to developing environmental foresight: UH Foresight Program for exploring the future insight into future environmental challenges of any topic (Hines and Bishop 2013). The pri- and opportunities, and the ability to apply that mary focus for this horizon scanning project insight to prepare for a sustainable future involves the first two steps of the Framework (Bengston 2012). Broadly speaking, the objec- Foresight process: Framing the topic and its tives of the horizon scanning system are to boundaries and scanning to identify emerging find, collect, and analyze the signals of change, issues. Framing and scanning provide the and to identify emerging issues suggested by foundation for forecasting, depicted as the these signals that could affect forests and for- baseline and alternative futures in Figure 1. estry in the future. This project will also use The baseline future or “business-as-usual” this information to support the development of assumes continuity with the present without scenarios of the future of forestry that integrate major surprises: trends stay on track, plans are signals of change and emerging issues into fulfilled, and mainstream projections are on each scenario. Once the scenarios are crafted, target. Emerging issues, however, may indi- indicators based on signals of change for each cate potential alternative futures, that is, alter- scenario will be identified. Going forward, the native outcomes to the baseline. Thus, the horizon scanning system can be used to moni- identification of emerging issues provides tor these indicators and provide early warnings early warning for potential shifts or disconti- that the future seems to be moving toward a nuities from business-as-usual and helps frame particular scenario (Schwartz 1996). This alternative future scenarios. information can alert decision-makers to adjust plans accordingly and take timely action where Framing necessary. In addition, the horizon scanning system is The process begins with framing the domain or supported by volunteers from within the Forest topic to be explored. The goal is to scope the Service. By including participants from topic so that it is neither too broad nor too nar- throughout the Forest Service, the project row (Hines and Bishop 2015, 374). For this seeks to foster a culture of foresight within the project, natural resources and the environment organization, and eventually to develop a more were considered as a broader scoping and for- forward looking organizational structure for ests as narrower. After doing some preliminary the USFS and other natural resource manage- scanning, it was decided that forests would ment agencies. provide more focus to scanners. Natural The next section of this article explains the resources-related scanning hits (e.g., energy, approach taken to develop the Forest Futures water) were too broad, but they could still be horizon scanning system. This is followed a included as they related to forests. Thus, scan- summary of what has been learned so far, and ners primarily focused on forests but were next steps for the project. open to how other natural resources might affect forests. This provided a more useful and Methods for Setting Up a focused framing than the future of natural Horizon Scanning System resources with a secondary focus on forests. The Forest Service partnered with the UH’s Domain mapping. The domain map is a visual Foresight Program to design and implement representation of the boundaries and key cate- the horizon scanning system, driven by a small gories to be explored, or framed, in scanning. core team with members from both organiza- Simple diagrams can be used to represent key tions. The Foresight Program trains students in categories and subcategories. Mind mapping the concepts, theories, and frameworks for software works well for this. We used the understanding, forecasting, and influencing cloud-based Coggle tool . the future (Hines 2014). A core approach called Most domains can be usefully mapped with Framework Foresight was developed in the three to six main categories and two to four Hines et al. 139

Figure 1. Key framework foresight elements for scanning project. subcategories of each of those. Typically maps •• What are the key activities that take also include a third level (subcategories of the place in the domain? subcategories). Beyond that, the map becomes •• Who are the key stakeholders in the messy as a communications tool, although the domain? greater level of detail in a domain map can be •• What has been driving change in the helpful to scanners. domain? A domain map has three primary functions: defining the boundaries of the scanning world, In some cases, one of these questions will organizing the data for analysis, and as a com- prove fruitful as an organizing theme. In the munication device for scanners. Detailed USFS case, a hybrid of all three proved useful. domain maps are helpful for those setting up Six first-level categories formed the core of the and managing the scanning process. But for map. Twenty second-level categories were linked most scanners, especially in a volunteer capac- to them. Third- and fourth-level categories were ity, a more streamlined/simplified domain map identified as appropriate, resulting in nearly one is more instructive and functional. hundred categories in total. Each of the primary The process of creating the domain map is categories is represented as a main branch: an iterative one. It starts with identifying the ecosystem, industry, institutions, stewardship, important categories and subcategories for climate, and STEEP (Social, Technological, understanding a particular domain. If the topic Economic, Environmental, and Political) (see is a new one, this can be challenging, and it Figure 2). The ecosystems branch could not only may require some preliminary background be viewed as a stakeholder but is also the context research to get a basic grounding in the topic. for most USFS activities. The industry branch The client or end user of the horizon scanning represents a key external stakeholder. The institu- system should be involved early in the process tions branch focuses on internal stakeholders. of defining the domain because they know The project team debated about whether to their domain well. This was the case for the include this in a domain map for horizon scan- USFS employees who were involved in this ning because these internal categories are not project. ones that are scanned for. Since the institutions Important questions to help identify key branch represents, in part, the client, it was topics in a domain map include the following: decided to keep it in to remind scanners of whom 140 World Futures Review 10(2)

Figure 2. Forestry domain map. the scanning was ultimately being done for. The done together. In this project, the Houston stewardship branch is the most focused on key team started with preliminary first-level cate- management activities of the Forest Service. The gories and a few subcategories, and assigned early iterations of the map included a branch of four student scanners to different pieces of the “issues,” dominated by climate change. After the map for preliminary scanning to see how fruit- preliminary scanning revealed the huge impact ful, or not, the categories proved to be. Each of climate change, it was given its own branch, student scanner was given one of four initial and the more general “issues” were dropped. It principal categories to scan for, and everyone was also judged by the core team that a general was responsible for scanning for STEEP as a category of “issues” was not helpful to guide secondary responsibility. We did not actively scanning because the identification of specific scan for the institution’s category in the pri- emerging issues was a key deliverable of the mary domain map pilot testing. project. Finally, the broader context within which Our experience is that this preliminary scan forests were contained was represented by using to test the categories can be done in about a the standard STEEP categories. This broader half-day. The team then reconvenes to revise context was important to be represented on the and create a second version of the map. This domain map as a reminder of the context for was then shared with the USFS team for their emerging issues that could affect forests and for- input and feedback, and was revised again as estry. In some domain maps, the STEEP catego- needed. A couple of iterations of revisions ries are represented as an outer circle around the is enough to get started. Minor tweaking of other scanning categories to represent the broad the map can happen throughout the project, milieu for forests. and this was our experience in this project— The Houston team drafted a first iteration to several small changes to the map were made share with the USFS team, but this can also be along the way. Hines et al. 141

Figure 3. Tagging domain map.

The detailed domain map can be intimidat- project in Sweden (http://www.vireoadvisors ing to those not involved in creating it. .com/blog/2017/3/14/swedish-processum-to- Scanners often add to it as they become more lead-major-nanocellulose-project) indicates comfortable with it. A greater number of sub- growing research activity related to wood- categories is helpful to them in providing based nanomaterials outside of the United greater guidance for the scanning. But for States, which could affect forests and forest those new to the project—in our case new vol- management in the United States. unteer scanners from the USFS—the map is likely to be seen as too busy and complicated. Timeframe. Forestry is a challenge to commu- Thus, we created a simplified version (see nicate to “outsider” scanners from the perspec- Figure 3) for inclusion in the Scanner Guide tive of the breadth and timeframe of the field. (to be described ahead) as a means to assign Forestry management often looks out one hun- consistent tags to emerging issues. dred years into the future, due to the nature of forest ecosystems. However, technology and Geography. Geographically, the focus was on industry change rapidly in the emerging wood the United States, but it was clearly within and fibers tech sector, especially in nano-bio- scope to scan for developments in other regions technology. This means that time horizons for as they might be relevant. For example, a scan- certain areas of our domain could look out ning hit describing a major nanocellulose hundreds of years, while other areas of the 142 World Futures Review 10(2) domain could experience significant change Guiding question. A guiding question captures within a year (see Bengston et al. 2017). At the why the topic is being investigated. Frame- same time, the Forest Service is influenced by work Foresight suggests there are two useful the regular short-term nature of budgets and types of guiding questions: strategic and elections that figure into any government exploratory. A strategic question guides a proj- agency. Therefore, domains need to be under- ect motivated by a specific purpose. For stood as multifaceted eras. For practical pur- instance, “should we invest in Blockchain poses, we used 2030 as the primary time technology?” The project is then designed to horizon. There was some subsequent learning provide insight to help answer the question. An on time horizon that we will discuss later. exploratory project, on the other hand, does not have a specific purpose, and the guiding Stakeholder analysis. Another important aspect question is more open-ended and aimed at of framing is to identify stakeholders who learning what the key issues or questions are could be interested in using the information, for a broad topical area. Our project was and who may have some influence over the exploratory, and the guiding question was, project or power to make decisions based on “What emerging issues might affect forests, foresight produced. Key internal and external forestry, and the USFS in the future?” While stakeholders for the horizon scanning project the horizon scanning and guiding question for were identified through discussions with the identifying emerging issues is exploratory, the Forest Service team. Likely internal (USFS) project also had the goal of setting up a formal stakeholders included the Forest Service ongoing horizon scanning system. Chief’s Office, USFS Washington Office lead- Framing sets the stage for the next step: ership, USFS Regional Foresters, National scanning. Forest and Regional Office planners, USFS Research Station leaders, and the Strategic Scanning Process Foresight Group itself. External stakeholders included state foresters, wood industry associ- Horizon scanning has sometimes been criti- ations, environmental NGOs, forestry societies cized for a lack of rigor, and most good scan- and organizations, forestry academics and sci- ners have difficulty communicating their entists, international forestry organizations, process for scanning (Hines 2003). Scanning, and the foresight community. and futures research in general (Burns 2005), A stakeholder map was created by position- is sometimes viewed as more art than science. ing each stakeholder group on a 2 × 2 grid Horizon scanning is often characterized more along the dimensions of their likely degree of by informal guidelines than methodological interest in the information produced by the rigor, especially when compared with the horizon scanning system and their perceived many variations of forecasting techniques. power to make decisions or policy based on the A “Scanner Guide” can bring internal guid- information. The resulting stakeholder map ance to support rigor and ensure consistency to can be used to identify individuals or groups to the scanning process (Figure 4). The guide is contribute to the horizon scanning project as aimed specifically at the goal of establishing volunteer scanners or to help in the analysis of an ongoing (internally sustainable) scanning the emerging issues. In addition, the stake- system. The chief benefit of a Scanner Guide is holder analysis can help the project team to assist in building the volunteer scanning develop multiple communication and dissemi- team. Horizon scanning is not part of anyone’s nation products tailored to key stakeholder official job description at the USFS and groups and an overall communication strategy. requires volunteer scanners. The project team This enhances the chances that stakeholders agreed that having a strategy for scanner will use the emerging issues in decision mak- engagement would be critical to the ultimate ing because the information is specifically success of establishing an ongoing scanning designed for their needs. system. The scanners will be drawn primarily Hines et al. 143

Figure 4. Scanner guide introduction and contents. from current USFS colleagues, former (retired) Another way to increase the rigor in scan- USFS employees, and for the first few years ning is to define a systematic scanning process. will also include Houston Foresight student Hines (2003) proposed a generic four-step interns. scanning framework, called FAFA: The Scanning Guide is envisioned as a living document that is continuously updated as •• Find: identify where and how to look new insights are gained into improving scan- for scanning hits ning effectiveness. The project is already on •• Analyze: use cross-level analysis and Version 9 of the guide, although many of the cross-layered analysis upgrades were fairly minor. The guide was •• Frame: develop a framework for orga- developed for both the Forest Service volun- nizing insights teers and the UH Foresight students. A chal- •• Apply: use the results in work processes lenge in developing the guide was to make it relevant to the diversity of scanners: The USFS The FAFA framework provided a starting team has more in-depth subject and technical point to further refine a methodology for scan- knowledge, while the Houston Foresight stu- ning projects. Figure 5 depicts the subsequent dents are already familiar with scanning. For scanning process that is currently being taught scanners associated with the Forest Service, by Houston Foresight (Hines and Bishop 2015, the forestry information in the Scanner Guide p. 381). This version suggests three principal serves only as a general refresher, while their steps in scanning: find, collect, and analyze. main focus will be on understanding what scanning is and how to effectively perform the Find. Find is the process of searching for and function. For Houston Foresight students, the identifying potential scanning hits. Scanning main focus is on learning more about the hits are new, unique, and potentially disruptive Forest Service and forestry, while the scanning ideas that could at some point have important materials is simply a refresher. impacts or become drivers of change or 144 World Futures Review 10(2)

Figure 5. Scanning process. emerging issues. The task of the scanners is to com). Diigo is easy to access and has a user- seek out these ideas and capture them. friendly interface, setup, and export functions The domain map categories from framing that are well-suited for a virtual team scanning provide a useful jumping-off point to organize project. the search. The categories on the domain map The purpose of collecting is to keep track of can be used as primary search terms, accompa- the scanning hits that might provide the basis nied by futures-oriented terms, such as “future,” for identifying an emerging issue. As scanners “trends,” “issues,” “long-term,” “change,” find an article, blog post, video, or whatever “vision,” “2030,” and so on. Getting the right item they would like to collect as a scan hit, search terms is less important than it was in the they use a “Diigolet” icon installed on their past, because many search engines now work web browser to link it to the team library in well with natural language inputs. But having a Diigo. The Scanner Guide provides instruc- list of potential search terms is useful to help tions for scanners on how to set up their web beginner scanners get started. Many tools are browser and link to the Diigo account. For available for finding and monitoring up-to-the- each scanning hit, the scanner provides a short minute information, such as Internet feeds and summary of why they selected the article. This alerts, as well as sources beyond simple search can simply involve cutting and pasting a engines such as specialized databases. descriptive paragraph from the piece itself and/ or can include commentary from the scanner. Collect. Collect is the process of storing and The scanner also adds a sentence or two about categorizing scanning hits after they have been potential implications of the scanning hit for identified. There are several online cloud- the project, in our case the implications of the based bookmarking tools with tagging capa- scanning hit for forests, forestry, and the USFS. bilities that can handle group inputs. The It is crucial that scanners tag their scanning convenience and functionality of these sites posts with a set of descriptors. This keeps the over an old-fashioned spreadsheet list and tags scanning library organized and easily search- are compelling. First and foremost, a geo- able. Framework Foresight uses the domain graphically diverse team can add their scan- map hierarchy as the basis of the tagging sys- ning hits to a private project library whenever tem. For example, if a scanner finds an innova- and wherever they are. A spreadsheet can be tive new use for a paper product, they would used in a cloud-based file-sharing system as tag it with “Industry,” “Forest Products,” and well, but it takes far more time and runs the “Paper.” This is not an exact science, but more risk of version control problems. The Diigo precise tagging aligned with the domain map site was used in the USFS project (www.diigo. leads to more efficient searching of the library Hines et al. 145 of scanning hits and aids in the analysis and a fundamental uncertainty, and the hit communication of results. The tagging system provides evidence for one of the possible enables a visitor to the library to quickly alternatives. For example, a scanning hit access, for example, all the ecosystem-related providing evidence of a paradigm shift in articles. The library’s front page keeps track of fire management from a “war on fire” to the top ten tags, which can provide an indica- “living with fire.” tion of whether certain topics are being •• A “5” or high score is assigned to scan- neglected or overemphasized. ning hits that suggest a “novel” future A tagging system based on the domain map possibility and have enough plausibility is useful in organizing the scanning library to be worthy of further consideration. (Houston Foresight 2014). For the USFS proj- For example, a scanning hit describing ect, the first- and second-level domain map genetic engineering to reduce the categories were to be used at a minimum as impacts of forest fires by making trees tags. Third- or even fourth-level tags could be less flammable. included, and it was suggested to include a few article-specific tags as well if necessary. The triage analysis can be used in several Scanning is an iterative process, and there is different ways depending on the goals of the flexibility to add new tags or even edit the map analysis. It could eliminate scanning hits from as the scanners learn more about the topic and analysis that were scored 1 if confirmational emerging issues. scanning hits were not important for decision makers to weigh possible future policy direc- Analyze. Analyze is a sense-making activity tions. In addition, the triage analysis could just that involves prioritizing the various scanning select the 5s if the goal is to provide emerging hits collected. Framework Foresight suggests issues on novel emerging issues. There may three degrees or levels of analysis, ranging also be a reason to tweak the scores in a par- from a simple triage to multicriteria rankings ticular project. For example, if the decision to sophisticated weighted indices. Some hori- maker or client is most interested in more plau- zon scanning efforts include pruning scanning sible and less speculative futures, the resolving hits that are deemed less relevant. This is effec- hits may be scored higher than the novel hits. tive when the focus of the horizon scanning The second level of analysis evaluates the effort is more targeted. In our case, all scan hits scanning hits that made it through triage and were kept in the data base. are further filtered using seven criteria: credi- The triage level of analysis involves mak- bility, novelty, likelihood, impact, relevance, ing a quick judgment about a scanning hit. time to awareness (Timeliness 1), and time to Framework Foresight uses a simple three-level prepare (Timeliness 2). Two or three criteria ranking system: from this list are often sufficient for narrowing down the scanning hits at this level of analysis. •• A “1” or low score is assigned to those There are several questions used for each of hits judged to be “confirming” what is the seven criteria to determine a scanning hit’s already fairly well-known. In our termi- level of each criteria. The questions are as nology, it confirms the baseline future. follows: For example, a scanning hit suggesting that wildfire management will consume Credibility a growing share of the USFS budget. •• Is the source reputable? •• A “3” or medium score is for those hits •• Are there confirmations elsewhere? that “resolve” in favor one of the major known alternative futures. It may be Novelty an issue in dispute, a driver that could •• Is the hit new? Or has it been widely play out in different directions, or reported? 146 World Futures Review 10(2)

•• Is it new to the client/audience? this could be a useful option. In addition, this analysis option could also serve to give more Likelihood weight to scanning hits with long or varied •• What are the chances that the hit will time horizons, which could be important for occur, and that it will amount to identifying emerging issues for forestry where something? the ultimate impacts to forests may happen decades or centuries into the future. Impact •• Will it change the future? Lessons Learned •• If it does change the future, how big a change will that be? This section describes what has been learned so far as the project enters its second year of Relevance operation. •• How important is that change to the client or the domain? 1. Background information versus •• Is the relevance direct or indirect? scanning

Timeliness 1 (time to awareness) The Framework Foresight process makes the •• How long before this information is distinction that background information or widely known? background research covers the recent history •• When will it appear in a mainstream and current conditions of the domain being newspaper or magazine? explored, and scanning covers what might be •• Are there resources to influence the changing in the future. Thus, scanning hits potential outcome suggested by the hit? should be relatively new in terms of when they were published—within the last few years is Timeliness 2 (time to prepare) our general rule of thumb. If something inter- •• How long before this hit begins to esting was said ten years ago, that is history change the future? and part of background information. In some •• Is it too late to do anything about it? cases, that “something said” was largely •• Is it so far off that action now would be ignored and thus appears as new information. premature? Our view is that it is still part of history and background research. For example, in the early Answers for each criterion will determine 1960s, Omer Stewart (2002) identified the role which scanning hits should be used in an anal- of fire as important for ecological systems ysis. As with triage, this is determined by the across the world and that fire has been used for goal of the analysis. For example, if the goal is thousands of years by indigenous people. This to find novel scanning hits from credible work was largely ignored by foresters and sources that take a long time to prepare for, ecologists for decades. Now, however, there is those scanning hits can be identified and a major shift in ecological thinking that con- analyzed. forms to many of Stewart’s ideas. While these The third level of analysis is a weighted ideas seem new on the surface, they should be index. This can be done by using the seven cri- considered part of the historical background teria above and weighting criteria deemed and not part of future emerging trends. more important to the project. Then a total number can be calculated for each scanning 2. “New to me versus new to the world” hit, and this can be used to list the scanning hits in order of importance according to the This is similar to the point above but can weighted criteria. This is “over-kill” for most involve recent information. The problem of projects, but in a scanning project in which the “new to me versus new to the world” is fre- scanning hits themselves are the deliverable, quently encountered in teaching beginning Hines et al. 147 foresight students about scanning, and the use detection of gravitational waves to a horizon of websites such as “How Stuff Works” as a scanning effort about forestry). source of scanning hits should be discouraged. These are great sources of background infor- 4. Moving beyond forests and forestry mation and can seem like new ideas or trends, but they are really just new to the scanner. A challenge for outside scanners, and in framing Everything can seem new and interesting to the domain, was trying to get “beyond forests” or one who is exploring a topic for the first time. “beyond trees.” The Forest Service deals with But some of this may be “old hat” to those with many concerns affecting forests and forestry experience in the field. Thus, it is important to organizations, including climate change, wild- calibrate whether something that seems new life, outdoor recreation, water, grazing, urban really is new. Having experts in forestry from forestry, indigenous rights, and many more. And the USFS involved was important in identify- all of these concerns are affected by social, tech- ing forestry related hits that were not new to nological, economic, and political change. For the agency or the field of forestry but may instance, the Scanner Guide suggested that scan- seem new to student scanners. Ecosystem ners “focus” mostly (but not entirely) on “out- management or ecological forestry, for exam- side” issues and change, that is, things that are ple, seem to be new concepts to nonforestry originating outside of the field of forestry and professionals; however, they are concepts with natural resources but could affect the field in the decades-old roots and far from novel within future. Many leaders and policy makers within forestry. the field are already aware of emerging issues and change originating within the sector. This 3. How to handle “coaching” of issue inspired a special project to develop a list of volunteers sources for the scanners to start with.

Some volunteers may not read the Scanner 5. Staying connected Guide and just plunge in and add hits that are off-track or below standard. Coaching and This is the opposite of the previous issue. other reminders about the goals of the scan- Some scanning hits seemed to be entirely dis- ning project can help keep scanners focused on connected from the concerns of forestry. useful hits. Our approach was to be careful to Granted, there was an explicit goal of connect- avoid being perceived as condescending or ing the external world to the Forest Service, overly academic in giving feedback to volun- but there did need to be some connection. The teer scanners. If the feedback is seen as too suggestion here was to ask scanners to add a harsh, the volunteers may become discouraged comment after the description of their scan- and drop out. Instead, we conducted team ning hit explaining its possible implications or “check-ins” to provide scanning tips. For relevance to forestry or the Forest Service. For instance, the issue of background information example, a possible implication for forests and being tagged as new scanning hits (see #1 forest management of self-driving cars is that above) prompted the suggestion to focus on their adoption could encourage more sprawl- recent emerging issues and developments— ing development patterns (as long commutes within the past year or so—rather than things are no longer wasted time), resulting in that happened years ago. Other ways that scan- increased fragmentation of forests. ners can go off-track are either being too focused on the present, so the hits proposed are 6. Stretching into the future not sufficiently future oriented (e.g., entering an article about ongoing deforestation in the The project team also sought to find a way to tropics), or the scanner entering hits that are encourage scanners to get further into the potentially game changing but for a different future. Scanners were asked to tag each of their domain (e.g., entering an article about the hits with the appropriate horizon (Figure 6): 148 World Futures Review 10(2)

Figure 6. Tagging on the three horizons. Source. Curry and Hodgson (2008).

•• Horizon 1: focuses on the current pre- Horizon 3 hits, but that will be monitored vailing system—the baseline—as it going forward. continues into the future, which loses “fit” over time as its external environ- 7. Tagging discipline ment changes. •• Horizon 2: an intermediate space of Tagging “discipline”—that is, accuracy and transition in which alternative futures completeness in assigning descriptive tags to begin emerging as the first and third scanning hits—can be a challenge. The tagging horizons collide. instructions in the Scanner Guide reminded scan- •• Horizon 3: focuses on “weak signals” ners to refer back to the domain map: “Tags about the future of the system, which should be 1st level [of the domain map], 2nd may seem marginal in the present, but level, 3rd level, something specific to the piece, which could signal significant change in and then which time horizon the hit targets.” the long-term Addressing this level of feedback may be a bit much to ask volunteers. The Houston team has The judgment of which time horizon is occasionally performed tagging tune-ups and most appropriate for a scanning hit is subjec- edited the library of scanning hits. Tagging disci- tive, but the process of tagging hits with time pline will be increasingly important as the library horizons may encourage more long-term grows. As of this writing, there were already thinking and more Horizon 3 hits, that is, if more than seven hundred hits in the library, so someone sees all their hits seem to be in finding items of interest would be a challenge Horizon 1 or 2, they could make an adjustment without an accurate tagging system. in their scanning approach. At the time of this writing, the breakdown 8. Current issues of hits by time horizon is 42 percent Horizon 1, 38 percent Horizon 2, and 20 percent Horizon To properly frame emerging issues it is important 3. It is not surprising that there are fewer to first identify a list of existing or current issues Hines et al. 149 facing the USFS. There is no clear source with a we have now reached the time for more empha- formal list of issues for the agency. Therefore, the sis on synthesis and communicating results. USFS-UH team reviewed Forest Service strate- The team is working on developing a horizon gic plan (USDA Forest Service 2015) and other scanning newsletter and periodic reports. This planning documents, to identify current issues. A article, and other planned journal papers as list of thirteen widely recognized current issues well as conference presentations, will also were identified, including the growing effects of communicate results of this horizon scanning climate change, more frequent and intense wild- effort. fires, and increasing forest fragmentation due to Because of the long-term nature of forests development. This list of current issues was and forest ecosystems, the agency has always added to the Scanner Guide to help scanners had a long-term (Horizon 3) time perspective. focus on additional emerging issues identified Our intention is that scanning will broaden this through horizon scanning rather than well-known long-term perspective to include emerging current issues. issues and weak signals of change that are in the external environment and which could sig- Going Forward nificantly affect forestry in the future. This horizon scanning system is in the early 2. Options for organizational structure stages of implementation. Our plan is to continue to raise awareness about this system The primary goal for this project has been to get through ongoing communication with key pol- the horizon scanning system up and running and icy makers and decision makers in the agency, generating useful foresight for the organization. especially in the Washington, D.C. office. Our A complementary and longer-term goal is to hope is to develop one of more high-level integrate foresight more tightly into the organi- “champions” for this work who will spread zation’s culture, decision making, and policy awareness of the effort and help disseminate making. It is too early to tell what effect, if any, specific scanning products (e.g., scanning the effort has had. The current scanning system newsletter, blog posts, analyses of key emerg- is designed around a volunteer model. The team ing issues identified through scanning, etc.). acknowledges that this could be a challenging The challenges identified in the preceding model to rely on for the long term. Volunteers section should provide plenty of work for the have many things competing for their time. project team to keep the horizon scanning sys- Once a proven concept, foresight may garner tem in good working order—current, relevant, support to continue as a staffed, dedicated effort. and consistent. There is plenty of additional The project might explore different options in work to communicate horizon scanning hits terms of developing ways to embed foresight and to develop information to feed into addi- into the organization. tional strategic foresight projects, like the inte- gration of the horizon scanning system with 3. Transitioning the horizon scanning sys- scenario development. Beyond what has tem to the Forest Service already been identified, there are a few other opportunities that the project team would like The joint project has the ultimate goal of devel- to explore. oping a system that can be run without assis- tance from the Houston Foresight program. It 1. Strategies for communicating results may be that Houston Foresight students could continue as interns, but the goal remains that As is to be expected in the beginning phase of the system be self-sustaining. An assessment a scanning project, a significant amount of of the system to determine its effectiveness information has been gathered. The Diigo and ongoing needs may be needed as the tran- scanning library is large and growing. sition to full operation by the Forest Service Synthesis always lags behind gathering, but approaches. 150 World Futures Review 10(2)

Conclusion NRS-121. Newtown Square: U.S. Department of Agriculture, Forest Service, Northern The USFS-UH horizon scanning project has Research Station. http://www.treesearch.fs.fed. provided an opportunity to experiment real- us/pubs/44822. time with academic approaches and in-the- Burns, A. 2005. “Is Futures Studies a Science or field practice of strategic foresight methods. An Art?” In The Knowledge Base of Futures Horizon scanning has often proven elusive to Studies: Professional Edition, Volume 1: teach and to institutionalize within organiza- Foundations [CD-ROM edition], edited by. tions. The project team has used a learning, Richard Slaughter. Indooroopilly: Foresight iterative approach to develop the scanning pro- International. Choo, C. W. 2002. Information Management for the cess that we hope will be sustainable within the Intelligent Organization: The Art of Scanning organization beyond the initial project. This the Environment. 3rd ed. Medford: Information article has described the set-up process and Today, American Society for Information what has been learned to date. The challenge Science. ahead is for the process to produce useful Curry, A., and A. Hodgson. 2008. “Seeing in results such that formal horizon scanning will Multiple Horizons: Connecting Futures to become an indispensable component of the Strategy.” Journal of Futures Studies 13 (1): work of the USFS as it moves into an increas- 1–20. http://www.jfs.tku.edu.tw/13-1/A01.pdf. ingly uncertain and challenging future. In de Haan, J. 2006. “How Emergence Arises.” addition to serving the needs of the USFS, it is Ecological Complexity 3 (4): 293–301. hoped that this project might also provide use- Douw, K., and H. Vondeling. 2006. “Selection of New Health Technologies for Assessment Aimed ful lessons for futurists and their clients in set- at Informing Decision Making: A Survey Among ting up systems in different fields. Horizon Scanning Systems.” International Journal of Technology Assessment in Health Declaration of Conflicting Interests Care 22 (2): 177–83. The author(s) declared no potential conflicts of Gunderson, L., and P. Longstaff, eds. 2010. “Managing interest with respect to the research, authorship, Surprises in Complex Systems: Multidisciplinary and/or publication of this article. Perspectives on Resilience.” Special feature. Ecology and Society 14 (1): 49. http://www.ecol ogyandsociety.org/vol14/iss1/art49/. Funding Hibbard, K. A., P. J. Crutzen, E. F. Lambin, D. The author(s) received no financial support for the M. Liverman, N. J. Mantua, J. R. McNeill, B. research, authorship, and/or publication of this Messerli, and W. Steffen. 2007. “Group Report article. Decadal-Scale Interactions of Humans and the Environment.” In Sustainability or Collapse? An Integrated History and Future of People on References Earth, edited by R. Costanza, L. J. Graumlich, Bengston, J. Peck, R. Olson, M. Barros, R. Birdsey, and W. Steffen, 341–75. Cambridge: MIT J. C. Leyva Reyes, and C. Malouin. 2017. Press. “North American Forest Futures 2016-2090: Hines. 2009. “How Accurate Are Your Forecasts? Aspirational Futures for Building a More More Accurate Than You Might Think.” World Resilient Forest Sector.” Review draft, USDA Future Review 5–22. Forest Service, Northern Research Station, St. Hines, A. 2003. “Applying Integral Futures to Paul. Environmental Scanning.” Futures Research Bengston, D. N. 2012. Environmental Futures Quarterly 19 (4): 49–62. Research: Experiences, Approaches, and Hines, A. 2014. “A Training Ground for Professional Opportunities. General Technical Report NRS- Futurists.” The Futurist 48 (5): 43. P-107. Newtown Square: U.S. Department of Hines, A., and P. Bishop. 2013. “Framework Agriculture, Forest Service, Northern Research Foresight: Exploring Futures the Houston Station. Way.” Futures 51:31–49. Bengston, D. N. 2013. Horizon Scanning for Hines, A., and P. Bishop. 2015. Thinking about the Environmental Foresight: A Review of Issues Future: Guidelines for Strategic Foresight. 2nd and Approaches. General Technical Report ed. Houston: Hinesight. Hines et al. 151

Houston Foresight. 2014. “The Future of Student U.S. Army Environmental Policy Institute. n.d. Needs: 2025 and Beyond.” Lumina Foundation, “Homepage.” http://www.aepi.army.mil/. Indianapolis. USDA Forest Service. 2015. “USDA Forest Service Millennium Ecosystem Assessment. 2005. Strategic Plan: FY 2015–2020.” FS-1045, U.S. Ecosystems and Human Well-Being: Synthesis. Department of Agriculture, Washington, DC, Washington, DC: Island Press. June. Munck, R. and G. McConnell. 2009. “University Strategic Planning and the Foresight/Futures Author Biographies Approach: An Irish Case Study.” Planning for Andy Hines is an assistant professor and program Higher Education 38 (1): 31–40. coordinator for the University of Houston’s Olson, R. D., and D. Rejeski, eds. 2005. Graduate Program in Foresight and is also speaking, Environmentalism and the Technologies workshopping, and consulting through his firm of Tomorrow: Shaping the Next Industrial Hinesight. His twenty-five plus years of profes- Revolution. Washington, DC: Island Press. sional futurist experience includes a decade’s expe- Pilkey, O. H., and L. Pilkey-Jarvis. 2007. Useless rience working inside first the Kellogg Company Arithmetic: Why Environmental Scientists and later Dow Chemical, and consulting work with Can’t Predict the Future. New York: Columbia Coates & Jarratt, Inc., and Social Technologies/ University Press. Innovaro. Sarewitz, D., R. A. Pielke Jr., and R. Byerly Jr., eds. 2000. Prediction: Science, Decision Making, David N. Bengston is an environmental futurist and and the Future of Nature. Washington, DC: social scientist with the U.S. Forest Service, Island Press. Northern Research Station, Strategic Foresight Schwartz, P. 1996. The Art of the Long View: Group in St. Paul, Minnesota. He is also an adjunct Planning for the Future in an Uncertain World. professor in the Department of Forest Resources New York: Currency Doubleday. and the Conservation Biology Graduate Program at Stewart, O. C. 2002. Forgotten Fires: Native the University of Minnesota, USA. Americans and the Transient Wilderness. Michael J. Dockry is a research forester with the Edited by H. T. Lewis and K. Anderson, U.S. Forest Service Northern Research Station, Norman: University of Oklahoma Press. Strategic Foresight Research Group, an adjunct Sutherland, W. J., M. J. Bailey, I. P. Bainbridge, assistant professor at the University of Minnesota, T. Brereton, J. T. A. Dick, J. Drewitt, and N. and an associate editor of the Journal of Forestry. K. Dulvy, et al. 2008. “Future Novel Threats He earned a BS in forest science and a PhD in for- and Opportunities Facing UK Biodiversity estry from the University of Wisconsin-Madison Identified by Horizon Scanning.” Journal of and an MS in forest resources from the Pennsylvania Applied Ecology 45 (3): 821–33. State University, USA. Sutherland, W. J., Mick Clout, Michael Depledge, Lynn V. Dicks, Jason Dinsdale, Abigail C. Adam Cowart holds a BFA and MFA in creative Entwistle, Erica Fleishman, et al. 2015. “A writing, an MBA, an MS in Foresight, and a Six Sigma Horizon Scan of Global Conservation Issues Black Belt. He is a writer as well as a planning man- for 2015.” Trends in Ecology & Evolution 30 ager at the largest retailer in Canada, where he focuses (1): 17–24. on developing programs and processes to improve the Sutherland, W. J., and H. J. Woodroof. 2009. “The future-oriented focus of the management teams across Need for Environmental Horizon Scanning.” the country. He is deeply interested in applying narra- Trends in Ecology & Evolution 24 (10): 523–27. tive techniques to foresight development.