A survey was administered electronically to informants prior to interviewing and was divided into two sections. One section asked informants to identify whether the resilience enhancement strategies that seaports identified through their RPIs were subsequently implemented (addressing RQ2). The other section gauged the resilience outcomes of RPIs for seaports by presenting informants with the following 10 adaptive capacity indicators (see further discussion in Appendix D) obtained from the academic literature and asking informants to rate their strength prior to and after the completion of their seaports' RPIs (addressing RQ3):

The change is subjective and qualitative and is based on informants' perceptions of risk and the value of assessments in enhancing resilience. Respondents rated capacities using a Likert scale from 1 to 5 with the options of Not Present, Weak, Moderate, Strong, and Very Strong.

We held 12 Zoom interviews of roughly 45 min each, nine of which were in focus groups of two to four individuals, and three were held individually with informants of the same seaport that could not participate together due to scheduling conflicts. We chose the focus group approach over individual interviewing where possible because the use of “mini-focus groups” (i.e., groups of three or four informants) is considered advantageous when participants have specialized knowledge and/or experiences to discuss in the group (Krueger, 2014)—we felt such was the case for our sample of seaports. Further, qualitative methods literature has suggested that the internet-based focus group format overcomes some of the methodological issues arising from face-to-face focus groups. For example, Walston and Lissitz (2000) suggest that group members that participate in virtual settings are less likely to withhold viewpoints that they perceive as embarrassing or inflammatory; and are more likely to express dissent with group members of higher positions, which is frequently cited as a shortcoming of the in-person focus group approach.

All interviews were recorded and transcribed, which was made known to informants prior to interviewing. The interview instrument was divided into four sections. The first section consisted of introductory questions designed to better acquaint the researchers with the seaport's resilience planning process, such as “what drove your organization to undertake an RPI?”. The second section focused on informants' perceptions of the key benefits or utilities of their seaport's RPI, either those associated with the process itself or the findings documented in the final report. The third section addressed challenges that the organization experienced along the course of the process and any aspects of the effort that were of limited utility. The final section focused on institutional impacts, such as whether it changed the organization's climate change planning culture.

We received survey responses from 19 of the 26 informants. All survey data was compiled in IBM SPSS Statistics (Version 26) and analyzed in aggregate. We categorized the strategies prescribed by each seaport's resilience planning report using six seaport-specific strategy typologies previously identified by Becker and Caldwell (2015):

Researchers tallied the total number of reported strategies belonging to each typology, along with the total numbers of strategies that informants indicated had been or will be implemented, may be implemented, were not implemented, and those that they were unsure about. We also asked informants whether a given strategy was likely or unlikely to have been identified and implemented in the absence of the RPI, as this would allow for further evaluation of resilience planning. Lastly, data from the second section of the survey—the strengths of the 10 indicators before and after RPIs were completed—were averaged across the 19 responses (see Appendix D for further information).

The research team coded interview transcripts line-by-line using the Atlas.ti v9 (Friese, 2019) qualitative data analysis software package to identify emerging themes regarding the useful and challenging aspects of undertaking resilience planning. The coding scheme used an iterative process based on grounded theory (Charmaz, 2006). Statements characterized as having a positive impact on or utility value to the seaport were coded as benefits of the RPI; statements evoking sentiments of difficulty or limitations were coded as challenges; and statements involving changes in the seaports' resilience planning culture (e.g., resilience becomes a budget item in a capital improvement plan) were coded as resilience enhancement strategies. This process allowed for views and concepts to emerge and be grouped into unique categories. Researchers recorded the number of times a benefit, challenge, or strategy type was mentioned and the number of seaports to which interviewees that mentioned it belonged. All benefits, challenges, or strategies, even those only mentioned once, were considered as findings in this paper. To ensure reliability of the coding scheme (see Appendix E for further information), we held two exercises in which several third-party individuals independently coded selected passages of transcripts using our coding scheme. We then calculated intercoder agreement using the Krippendorf's Cu-alpha/cu-alpha coefficient, yielding a value of 0.796, suggesting very good agreement (Friese, 2019).

Key informants from nine of the 10 case studies described a more comprehensive and nuanced understanding of their seaports' vulnerabilities as a benefit of their RPIs (24 mentions voiced by 14 informants). In many instances, informants described their RPI as their seaports' first detailed investigation of their vulnerabilities. The impetuses to undertake these efforts were either a state mandate (as was the case for four case studies, though some started resilience planning voluntarily prior); to obtain federal mitigation funding (two case studies); growing recognition of the threats posed by climate change as evidenced by recent natural hazard events (one case study); or for some other reason (three case studies). Interestingly, some informants felt that, along with identifying risks to proactively mitigate, their RPI revealed what not to worry about. Several informants also valued the byproducts of their planning, such as inventories of their seaport's vulnerable assets or GIS-based inundation maps, as this information allowed them to better understand the geographic extent of climate risk and aided their roles in their respective departments.

The enhanced quality of vulnerability information was most frequently emphasized as a benefit, as one of the fundamental objectives of resilience planning is to elucidate this. It is also intuitive that informants would value a process that yields information about the localized risk to their seaports' future capital improvements, as this information enhances their organizations' capacities to satisfy their civic responsibilities of economic development and facilitating trade. Previous research on seaport decision makers' perceptions of climate change (Becker et al., 2012) and barriers to planning for it (Mclean and Becker, 2019) may further explain why this benefit was most frequently mentioned. Often, nuanced vulnerability information is lacking among seaports (Becker et al., 2012; Mclean and Becker, 2019), serving as a broader barrier to the process of building resilience and adapting (Moser et al., 2010). For example, in a 2012 study from Becker et al., more than half of seaport respondents felt that they were not sufficiently “informed” about climate change impacts to their facilities. Years later, in interviews with 30 seaport decision makers, Mclean and Becker (2019) documented a widespread lack of understanding of climate and natural hazard risks among a majority of their interviewed seaports. Our results suggests that RPIs amend a widely recognized barrier to climate action at seaports.

Twelve informants, representing eight of the 10 participating seaports, found resilience planning to enhance their seaports' social capital with internal and external stakeholders (24 mentions; 12 informants). Social capital defines the “resources embedded in relationships among actors” (Häuberer, 2011, p. 50), and can be thought of as a function of the quality, quantity, and direction (i.e., horizontal or vertical) of relationships of trust, reciprocity, and exchange within a system (Adger, 2003; Häuberer, 2011; Djalante et al., 2012). Informants remarked how the resilience planning process enhanced social capital both during and after planning was complete. Many of the informants felt that engaging key stakeholders—especially during the preliminary planning phases of scoping and defining objectives—built social cohesion and facilitated mutual understandings amongst different departments and, in some instances, with the external community. One informant remarked how his seaport's inclusion of external stakeholders (city officials, NGOs, state government officials, etc.) provided a systems perspective of vulnerability otherwise unattainable through conventional planning approaches. In the opinion of another informant, his seaport became a centralized hazard information network for the surrounding community (DIR, January 2021). Interestingly, seaports' improved relationships with stakeholders were often byproducts of another interrelated benefit: the RPI served as a boundary object—a process, product, or other form that bridges communities, stakeholders, and disciplines and leads to links from knowledge to action (Star, 2010)—that stimulated dialogue amongst departments that often do not communicate with one another.

Our findings are consistent with previous research of seaport stakeholders' perceptions of collaboration in seaport resilience efforts. Becker (2017), for example, convened seaport stakeholders in workshops based on three boundary object scenarios, revealing the utility of boundary objects as planning tools that facilitate critical thinking amongst diverse stakeholder groups to create a shared mental model of risk. Similarly, in interviews with 25 stakeholders of the Port of Providence, Rhode Island, Becker and Kretsch (2019) provide empirical results capturing perceptions of the value of collaborative efforts as necessary to implementing resilience strategies.

Another benefit that was voiced in six of the 10 focus groups was the impact that resilience planning had on seaports' leadership (12 mentions; six informants). Decision makers explained how, prior to their seaports' efforts, their leadership did not view resilience as a pressing matter that warranted capital expenditure, stymying long-term resilience-building efforts. When reports from the interventions were presented to lead decision makers, however, informants felt that leaders gained a heightened awareness of the exigence for resilience-building across the system, as suggested by Wilkinson (2012b).

The finding that RPIs may result in a more resilience-conscious and motivated leadership may correspond with the aforementioned benefit of enhanced understanding of vulnerabilities, although this assertion warrants additional study. The information products that come out of resilience planning during the assessment phase, such as SLR inundation maps, may add tangibility to impending threats posed by climate change (Retchless, 2018; van Valkengoed and Steg, 2019) for leaders who are otherwise preoccupied with the short-term concerns of running a public enterprise. The leadership benefit is particularly significant given the key roles leaders play in the resilience-building process, such as incorporating resilience considerations into policies and budgets, using deadlines to instill a sense of urgency amongst their organizations, providing information on new ideas, problems, and solutions to resilience partners, and building stakeholder networks (Stiller and Meijerink, 2015). Moreover, institutional voids in leadership for resilience-building constitute a major resilience-building barrier (Moser et al., 2010) that has been documented in seaport organizations (Becker and Kretsch, 2019; Mclean and Becker, 2019). Our findings suggest that resilience planning efforts may help transcend the obstacles presented by rigid administration, and, in doing so, enhance seaport adaptive capacity.

Five of the 10 seaports described their organizations' enhanced political efficacy in climate change conversations as a benefit of their RPIs (seven mentions; five respondents). In the opinion of several informants, resilience planning enhanced their organizations' abilities to engage in political arenas that previously challenged staff that were not accustomed to climate change jargon or concepts. Mentions of Benefit 5 coincided with mentions of another benefit: divisions of the seaport became more adept at funding resilience projects (11 mentions; five informants). Informants explained that the ability to sway decision-making in favor of resilience was enabled, at least in part, by the vulnerability assessment findings, which gave them data to better argue the value of public investments in resilience projects. In some cases, the ability to mobilize funds for resilience projects also improved. In the opinion of a director,

Both benefits have important implications for the role of using a resilience planning process in facilitating adaptive capacity. The finding that several organizations' political efficacy in climate change discourses improved, suggests that, because of the RPIs, decision makers become better poised to advocate for their seaports' resilience needs—a capacity that has been integral to the transformation toward more adaptive social-ecological systems (Menzel and Buchecker, 2013). By working with different departments and incorporating climate change expertise (e.g., from consultants), decision makers may learn how to “talk the talk” of climate change. The improved abilities to mobilize and advocate financial resources after the RPI also have direct implications for seaport adaptive capacity, given the unique funding that climate change resilience projects require (Smit and Wandel, 2006; Moser et al., 2019). This, in turn, may point to the significance of the discursive processes among interests associated with system as a whole (Innes and Booher, 2010).

Two additional benefits were mentioned in conjunction with one another. Four seaports valued how RPIs formalized seaports' strategic planning for climate change (eight mentions; four informants) and two felt that this motivated personnel from different divisions to champion resilience initiatives (four mentions; four informants). Informants explained that their seaports' climate change planning was largely an internal discussion with senior leadership or addressed by different departments in isolation, prior to their RPIs. Following an intervention, however, an engineer noted, “The vulnerability assessment standardized how we approach projects from a resiliency standpoint… It was kind of that ‘kindling for the fire' for our organization's resilience planning.” (ENG, September 2020). The formalization of climate change planning inspired staff of some seaports to carry out resilience projects in their respective departments. “At first, our engineering director was like, ‘We can't afford to go above and beyond building code. We're not going to add resilience.' But now, they are adding resilience into their projects and even applying for federal grants,” an environmental specialist explained (ES, November 2020).

Again, the significance of these benefits is apparent when compared with findings of Mclean and Becker (2019). Over half of the seaport decision makers in their study acknowledged a governance disconnect—describing the lack of coordination across sectors, or across levels of an organization, or both—as a key barrier to their seaports' efforts to build climate resilience. The governance disconnect barrier has been found to result from the lack of a management plan for climate and extreme weather adaptation (Moser et al., 2010). Benefit 7 clearly demonstrates the ability of resilience planning to fulfill this role. Additionally, governance disconnect may arise from institutional crowdedness, resulting in a lack of clarity of responsibilities for adaptation at local levels (Mukheibir et al., 2013; Ekstrom and Moser, 2014). Supported by Benefit 8, and consistent with resilience planning and adaptive management theory, it seemed that resilience planning helped formalize seaports' approaches to climate change. We posit that the multitude of seaport divisions may better understand their roles in their seaports' resilience-building efforts, leading them to actively champion the work that pertains to their respective departments (Schauppenlehner-Kloyber and Penker, 2016).

The most frequently mentioned challenge of RPIs involved engaging stakeholders (20 mentions; 10 informants). Our data reveal that stakeholder engagement was a challenge in all phases of the resilience planning process. For example, during the preliminary organization of the assessment, convening stakeholders was complicated by schedule conflicts or their views that the assessment was not worthy of their time. One safety planner remarked about the difficulty of conveying to stakeholders the value of participating in an exercise with no immediate or tangible benefits, as processes like disaster mitigation and prevention are “difficult to measure” (SP, November 2020). Additionally, RPIs necessitate discussion of vulnerabilities, often requiring participants to disclose sensitive information, which they may be reluctant to do.?? Scoping the assessment and defining objectives were also noted as bureaucratically cumbersome. For example, one informant mentioned the challenge of reaching consensus among his seaport's myriad stakeholder groups regarding the appropriate climate scenarios to plan for (ES, September 2020). Following the completion of their seaports' RPIs, several informants emphasized the challenge of communicating the vulnerability assessment findings to stakeholders and educating them about how to use the assessment (ES, September 2020). Two other focus groups' participants were challenged in their efforts to continue dialogue about the assessment after it was complete or raise awareness of the assessment to other departments that had not participated. In the opinion of informant,

The challenge of coordinating a multitude of actors with individual interests in resilience affairs is documented in seaport (Becker, 2017) and in the collaborative planning and governance literature (Margerum, 2011). Sellberg et al. (2015) report the convening of a diverse stakeholder set to participate in an urban planning resilience assessment, as a key challenge with resilience planning. In interviews with Australian natural resource management practitioners who undertook resilience planning, Sellberg et al. (2018) identified the involvement of different organizational departments in the development of a resilience plan, as a primary challenge of resilience planning. Similarly, interviews with city officials leading their cities' resilience plans underscore the difficulty of reaching consensus on the starting point for resilience-building and the definition of resilience to operationalize when multiple interests are engaged (Fastiggi et al., 2021). This challenge is even noticed as a barrier in the adaptation planning process in synthesis research investigating myriad adaptation contexts (Moser et al., 2010).

While the most commonly mentioned benefit was the enhanced vulnerability information, some informants acknowledged the limitations of the information their RPIs provided. Some seaports completed their RPIs over 5 years ago, when, as several informants mentioned, the science for certain climate hazards was less accurate and available as more recently. Informants from three focus groups felt that the lack of accurate, locally relevant climate hazard data (e.g., sea level rise projections) limited their seaports' abilities to identify and plan for those respective hazards (four mentions; four informants).

The finding that only three seaports mentioned this limitation is noteworthy. Though many seaports had completed their RPIs more recently, and thus had access to more accurate scientific information, this does not mean that uncertainties did not exist in their information products. For this reason, it was anticipated that this challenge would be more frequently mentioned. There are many propositions that could explain why more cases did not emphasize this challenge, though none can be proved empirically. For example, greater confidence in recent climate science (Reidmiller et al., 2018) may have provided those decision makers with more recently completed RPIs a greater sense of assurance in the vulnerability findings that came out of their assessment. Just as likely, due to the open-ended nature of the interview questions, other challenges may have been more palpable to the focus groups; hence, the absence of a mention of this challenge does not mean that it was not a challenge. Whatever the reason, this challenge suggests the administrative imperative of embedding flexibility in seaport policies and management to account for emerging information.

Several informants noted how the RPI that their seaport undertook was different than conventional planning procedures, for example, because of the larger time horizons considered or the integration of numerous stakeholder groups. Informants from two case studies expressed the difficulty of organizing a planning process with which they had little experience and that had no model to reference, as a challenge (three mentions; three informants). In the opinion of one inform,ant

Another informant from the same seaport explained that, unlike conventional risk assessment approaches, his seaport's RPI was necessarily improvised as it progressed. Unsurprisingly, when asked how they would execute their RPIs differently knowing what they do now, informants explained that they would seek advice from colleagues at other seaports that had already undertaken a similar effort. They had not anticipated the need to design the process to be inclusive of the multiple interests and authorities that were included as the process evolved.

An unanticipated challenge mentioned in two focus group interviews was communicating the vulnerability assessment results in a manner that would not harm the seaports' marketability to future lessees and investors (two mentions; two informants). Informants that mentioned this challenge felt that disclosing information about their seaports' vulnerabilities to external stakeholder groups may deter investment into their lands. For one informant's seaport,

As discussed earlier, U.S. port authorities and agencies act as “public enterprises” that have civic responsibilities while also competing to secure market share, market their services, and facilitate economic development via private enterprise (Fawcett, 2006). Therefore, decision makers that wish to undertake a RPI or similar initiative may want to include a communication strategy for navigating the potential publicity issues of disclosing vulnerabilities. Unlike the three previous challenges, we find this challenge to be less generalizable, as the public-private duality that characterizes many seaports distinguishes them from other entities (e.g., natural resource management agencies).

We counted 155 discrete strategies from eight of the 10 case studies' resilience planning documents¹ (and several others during interviews) (Figure 4). Of these 155, we found that construction and design strategies were most frequently mentioned (N_m = 60) and implemented (N_i = 25) after RPI completion; however, no statistically significant difference in quantities of implemented strategies existed between the six typologies (p = 0.689 > 0.05, Fisher's exact test, two-sided). Strategies falling under the construction and design typology consisted of developing and implementing physical changes either on or off the seaport. The most frequently mentioned strategy of this type was reinforcing structures, such as terminal assets, with more durable materials (N_m = 16); however, this strategy was also most frequently marked as unsure, owing to one case study's non-response to this section of the survey. The most frequently implemented strategy of the construction and design typology was stormwater management infrastructure improvements (N_i = 9). Following construction and design strategies, we identified 31 total emergency preparation, response, and recovery strategies; 28 research strategies; 18 networks and new ways of thinking strategies; 10 long range planning strategies; and eight building codes and land use regulations strategies. In terms of total quantities implemented, research strategies were the next most implemented (N_i = 12) after construction and design, followed by networks and new ways of thinking strategies (N_i = 11), emergency preparation, response, and recovery (N_i = 11), long range planning (N_i = 6), and, finally, building codes and land use regulation (N_i = 4).

Overall, the survey findings suggest that the main dimension of resilience that seaports prioritized in their adoption of prescribed strategies was infrastructural (i.e., technical resilience), more so than strategies addressing governance, management and strategy, and operations and production (i.e., organizational resilience). One potential explanation of why construction and design strategies were most frequently mentioned and implemented, is that infrastructure improvements and modifications are going to be pursued regardless of changing hazard risk levels. Without functional infrastructure, the seaport's capacity to facilitate the transfer of cargo is compromised; thus, having resilient infrastructure is merely complementary to the seaport's mission. Nonetheless, the insignificant difference between implemented resilience enhancement typologies may suggest that resilience enhancement strategies are too case-specific for cross-seaport comparisons.

Most respondents were unsure about whether their seaports might have identified or implemented specific strategies in the absence of their RPIs. We counted only the responses indicating that implementation of a given strategy was likely—suggesting that the RPI did not influence that area of the seaport's resilience portfolio—and not likely—suggesting that the RPI introduced the seaport to areas of resilience improvement. Overall, the informants found their RPIs to have the greatest influence on the implementation of monitoring systems that continually track environmental conditions (such as sea level height) or infrastructure damage, which fell under the research typology. By contrast, most respondents felt that participating in or establishing a climate change-related working group or ad-hoc committee was likely to be an implemented strategy in the absence of the intervention. The inability of most informants to indicate whether implementation was likely in the absence of the RPI, may indicate a weakness in the survey instrument to address the sought inquiry—the question may have been too speculative for informants.