Public priorities on locally-driven sea level rise planning on the East Coast of the United States

Survey responses

503 respondents completed the survey including the consent form in question 1 (Q1). 235 respondents reported living in a coastal community, 69 working in a coastal community, and 284 regularly visiting a coastal community (Q2). These add to more than 503 because some respondents had more than one affiliation. There were several significant differences between residents (235) and non-residents (268), as described in the priorities and vulnerability sections.

To provide context on the perceived importance of sea level rise relative to other issues, respondents were asked to state the importance of ten different key issues (Q3), including both environmental issues (such as preparing for sea level rise or protecting the environment) and non-environmental issues (such as reducing taxes or maintaining roads and other transportation infrastructure), on a Likert-type 1-5 response, with 1 meaning “very unimportant” and 5 meaning “very important”. For this and all future questions, the definition of every value in the Likert-type ranking, the full question, and each selection is available in the “questions and response meanings” in the supplemental materials (write-in responses are included in the full data set). This question was designed to assess the relative priority of sea level rise planning compared to both issues which are potentially close proxies for sea level rise planning in the public’s mind (“preparing for climate change” and “protecting against future flooding”) and those that may have some connection but are probably not directly linked in the public’s view (such as “helping people with limited resources” and “growing the economy”). Table 2 shows the responses to this question, showing that although “preparing for sea level rise” was one of the lowest ranked issues by both mean (3.68) and percent ranking as 4 or 5 (65.4%), other closely related issues such as “protecting the environment” (4.04 and 75.7%) and “protecting property from natural disasters (3.99 and 75.3%) were ranked more highly.

Survey participants were asked to comment on the importance of various potential components of a local sea level rise plan in a Likert-type 1-5 response (Q4), with 1 meaning “very unimportant” and 5 meaning “very important”. This question sought prioritization of sea level rise plan components recognizing that time and resources are likely to be limited. The responses to the question, which are detailed in Table 3, include highly ranked activities around preparing to respond when flooding happens, implementing required policies to mitigate future flood damage, and developing maps and tools to assist in planning. On the other end of the spectrum, fewer than half of participants ranked “finding ways to postpone making change until more research is done” highly, recognizing the need to start planning processes where they have not already begun. This question did not differentiate between postponing all action and developing plans where some decisions can deliberately and constructively be deferred to future under a real options approach (Dobes, 2008; Hoekstra & De Kok, 2008; Linquiti & Vonortas, 2012). This phenomenon could be explored in more detail during the development of SLR plans. Closely related to this question was the following one (Q5), which asked participants to write-in any additional SLR plan components, of which the most common response (20.8% of 120 coded responses) was education.

Respondents were asked about their perceived vulnerability to four natural hazards—water surge damage, repeated flooding from high tides, increased flooding from SLR, and other natural disasters (Q6) on a Likert-type question, with 1 meaning “not at all vulnerable” and 5 meaning “exceptionally vulnerable”. This question gauged how vulnerable respondents felt about these topics, rather than any objective measure of vulnerability (which would have required data not collected in this study). Figure 1 shows mean vulnerability for each component of this question (more detailed information is shown in Table S1). Overall, respondents found themselves to be the most vulnerable on average to damage from hurricanes and severe storms but also perceived some vulnerability from other hazards.

Respondents were presented with 15 potential protection priorities (which services and places to focus protection on with a SLR plan) (Q7). The intent of this question was to help identify what types of local services and amenities are considered by the public to be the highest priorities for inclusion in a sea level rise plan. These included a wide range of options, including various types utilities and related infrastructure (drinking water, electric power, sewer/wastewater and others), as well as individual homes, places of cultural importance, and others. These results are shown in Table 4. Although many essential services and others were highly ranked, drinking water was the only to exceed 80% of respondents ranking highly. Electric power, roads and highways, homes and residences, and sewer/wastewater were all at greater than 70% ranking as 4 or 5. When respondents were asked to identify other priorities not listed (Q8), those with more than five responses include medical facilities/hospitals (11 responses), educational facilities/school (10), and animal shelters/zoos (7). It is important to recognize that many services and amenities are ultimately interdependent within a community and may also rely upon actions within other communities, and therefore a prioritization is only a starting point for more in-depth local assessment.

To help better understand how the public perceives that local sea level rise plans can be developed, funded, and administered, respondents were asked about their preferences on whether the responsibility for preparing for future flooding and sea level rise should be the responsibility entirely of the public sector, entirely of the private sector, or somewhere between the two (Q9). The distribution of these responses is shown in Fig. 2. In this case, over 60% of respondents (303) selected “equal mix of public and private sectors” and of the remaining, more selected “mostly public sector” (115) than any other choice. This suggests that although the public generally looks more towards the public sector, the private sector is also viewed as a key contributor in SLR planning.

Recognizing that funding can be a significant challenge for implementation of any community-wide project, whether or not related to SLR, respondents were asked about the usefulness of various funding mechanisms in their communities (Q10), going from “not at all useful” (1) to “exceptionally useful” (5). Recognizing that the actual funding need and availability will vary considerably based upon other aspects of the SLR plan and other conditions within and outside of the community (such as availability of various funding programs), this question focused on the perception of how useful various funding mechanisms are, rather than the actual funding amounts or anticipated availability. Ten options were available, including voting on methods to pay for highest priorities, increasing various forms of taxes, and others. The summary of these responses can be seen in Table 5. Although no funding mechanism had greater than 60% of respondents rank it in one of the top two rankings, “hold public meetings to identify highest priorities and vote on methods to pay for them” was the closest, obtaining 59.2% of responses ranking as 4 or 5. On the opposite end, all forms of increased local taxes received a ranking of 4 or 5 by fewer than one third of respondents. Respondents were also given the opportunity to describe other mechanisms (Q11). Notable amongst these are responses that can be categorized as “governmental action/funding” (6), “improved information” (6), and “donations/fundraising” (6).

Recognizing that there are tradeoffs (including cost, complexity, and level of protection) to all hazard mitigation decisions, respondents were asked to indicate the desired level of protection (which could also be interpreted as tolerance for failure of those protections) from the cumulative protections of their SLR plan. Respondents were asked from ranges as frequent as failing less than 1 in every 10 years all the way up to failing less than one in every 1,000 years, and they were asked both about minor flooding and major flooding. Figure 3 shows the distribution of responses, including a general preference for failure less than 1 in 100 years for both major and minor flooding, although there considerable distribution across the choices, with protections generally desired to be stronger for addressing major flooding than for minor flooding (more detailed information is available in the Table S2). Although the question noted that protective measures that are less likely to fail are likely to be more expensive and complicated, there was no way to identify exactly how much so for the purposes of this survey. Rather, communities would need to study the specific tradeoffs amongst various protective measures as part of their planning processes.

Some of the decisions that will need to be made in developing and implementing a local sea level rise plan will probably be controversial. For this reason, respondents were asked to rate how helpful eight different methods to resolve conflict locally are likely to be, with 1 meaning “not at all helpful” and 5 being “exceptionally helpful” (Q13). Of the methods, the most favorable was considered to be discussions with preparedness experts about improving protection against floods. Discussions with scientists and increasing educational efforts were also high on the list. Table 6 shows these results. Respondents were also asked to write in any other methods that may be effective for resolving conflict (Q14). The most common responses coded to “community meetings” with eight, over 25% of all of the write-ins for this question.

The last two primary questions were around the perceived appropriateness of various adaptation responses that could be undertaken to increase resilience while adapting to SLR. The first question addressed these from a list (Q15) and the second allowed for additional write-in responses (Q16). In Q15, the responses ranged from “very inappropriate” (1) to “neither appropriate nor inappropriate” (3) and finally to “very appropriate” (5). Of a list of ten adaptation options, including a wide range of choices such as early warning tools, raising elevations on new construction (and/or existing construction), and hardening public infrastructure, all options were generally considered appropriate by having medians above three (neutral), except for “increase cost of insuring high-risk areas” with a median of 3, and “don’t provide assistance for areas at highest risk” with a median of 2. Table 7 shows these responses.

There was limited participation in identifying other potential adaptation measures (n = 21 respondents with a total of 26 responses). Of those that did respond, six coded to “improve public infrastructure” which is very similar to one of the responses in the previous question.

Part of the survey was the collection of a series of pieces of demographic information self-reported by each respondent. These included a household income range, self-rated level of environmentalism, job title, level of education, ethnicity, political affiliation, age range (Q24), gender (Q25), location/state (Q26). Summaries of these responses are included in Tables S3–S10. Respondents were also asked to provide any feedback of concerns about the survey (Q23), for which the most common responses (other than responses indicating no feedback) were either something positive about the survey (17) or concerns about sea level rise itself (11) or the wording of the survey (7).

The primary use of the demographic information was to analyze similarities and differences in priorities across demographics, presented in the statistical analysis below. Therefore, the demographic information is not presented in full here but is available in the supplemental materials. Overall, although some of the distributions are not perfectly representative of the underlying population (for example, female respondents are overrepresented compared to the underlying population), they are diverse enough to represent a substantial number of viewpoints.

Statistical analysis

There were several key differences between those respondents who are residents of coastal communities and those who are not (i.e., those who either worked in or regularly visited a coastal community but were not also residents). Of the 503 respondents, 235 lived in coastal communities and 268 did not. 65 potential differences (all of the subcategories of each primary question) were tested using Mann–Whitney U (nonparametric) tests. Five relationships which were found to be significant. The following five statements are provided first with the Mann–Whitney U p-value.

• Residents perceived their communities to be more vulnerable to hurricanes and severe storms (3.57) than non-residents perceived (3.16) the communities they worked in or regularly visited to be (p = .001), with a difference in mean of 0.191 to 0.628 (95% confidence interval).

• Residents perceived their vulnerability to repeated flooding from high tides (3.10) to be greater than non-residents (2.70) perceived the communities they were associated with ( p = .001), with a difference in mean of 0.161 to 0.631.

• Residents perceived their vulnerability to increased flooding if sea level rises in the future (3.40) as higher than non-residents (2.70) perceived the coastal communities they were affiliated with (p < .001), with a difference in mean of 0.217 to 0.667.

• Residents placed higher priority on the importance of electric power for sea level rise plans (4.34) than non-residents (4.13) placed on electric power (p = .016), with a difference in mean of 0.046 to 0.381.

• Beaches and similar coastal areas were given a higher priority by residents (3.91) than non-residents (3.62) (p = .008), with a difference in mean of 0.095 to 0.486.

Figure 4 shows these significant relationships. As there were five statistically significant differences between coastal residents and non-residents out of 65 relationships (7.7% of relationships), there are important differences between the two groups, especially with regards to perceived vulnerabilities (3 of 5 significant relationships) but the overall difference in priorities and preferences was modest.

Some demographics correlated with changes in the distribution of responses for some or all components of primary questions. Overall, the self-reported level of environmentalism predicted the largest number of changes to the distribution of primary question responses, with 62% of components to questions likely influenced by this demographic. The preferred funding mixture (public, private, or equal mix) was the second most powerful predictor, coming in at 32% of components to primary questions. Gender and Age were each 29%. Notably, ethnicity, education, and income all influenced less than 10% of question components, and political party only 12%. A summary of these results is shown in Table 8. The component tables showing the P value of every test on each component of each question are found in the supplemental materials.

Key findings summary

Six key findings were developed as described in the methods section. The justification for each finding is discussed within the following discussion section. The key findings were:

Finding 1 on relative priority: “Officials are likely to gain better engagement with the public if they make a strong connection between planning for sea level rise and other high priority issues like the environment, infrastructure/utilities, and the economy”. This finding is informed by the results to question 2 and Table 2.

Finding 2 on planning components: “Officials should consider building sea level rise plans that integrate response planning and preparedness with mandatory policies to reduce future damage. Maps and tools, educational resources, and voluntary protections were also popular, but inaction to wait for more research was not popular”. This finding is informed by the results to questions 3 and 4 and Table 3.

Finding 3 on protection priorities: “Officials should consider the protection of essential utility and transportation services as some of the highest priorities for protection in sea level rise plans. Residents also rate the protection of individual homes and of government facilities very highly”. This finding is informed by the results to questions 7 and 8 and Table 4.

Finding 4 on funding priorities: “Funding may be one of the largest challenges of sea level rise planning. Officials should consider public meetings to discuss how to pay for priorities, should use state and federal funds when available, and should work with the insurance industry on risk reduction measures. Officials should avoid cutting other programs and should proceed cautiously with taxes”. This finding is informed by the results to questions 10 and 11 and Table 5.

Finding 5 on conflict resolution: “To help prevent and resolve conflict, officials should consider bringing in both preparedness experts and scientists familiar with flooding and sea level rise to talk with the community and use the media to help educate the community about the issue. Avoid making adaptation measures optional to avoid conflict”. This finding is informed by the results to questions 13 and 14 and Table 6.

Finding 6 on adaptation responses: “Public officials should consider a variety of adaptation responses. Early warning systems, natural and artificial barriers, and hardening infrastructure are among the items respondents generally found to be appropriate. Even some potentially controversial adaptation, such as preventing new development in vulnerable areas were generally viewed as appropriate. Officials should avoid cutting off assistance from high risk areas”. This finding is informed by the results to questions 15 and 16 and Table 7.

Key findings discussion

The six key findings of this study were developed based upon the study’s findings and were written in plain language to be of maximum utility for public officials, as discussed above.

For relative priority (finding 1), respondents ranked preparation for sea level rise relatively low on the list of other issues (question 2, shown in Table 2), indicating that in many instances they may not fully engage unless they make connections to other issues that are higher priorities, such as the environment more generally or the economy. The long-term and somewhat abstract nature of sea level rise may put it in the back of people’s minds, they may not fully understand it, or other barriers may lead to challenges for public officials in engaging the public. SLR plans that link adaptation measures for SLR with other concerns such as maintaining and building economic opportunities, protecting other facets of the environment, and others may assist in gaining needed engagement. Additional study on this phenomenon could yield additional insights, as the reasoning for respondent’s answers is not known from this study alone.

For planning components (finding 2), a wide variety of components, such as response plans, mandatory mitigation policies, and maps and tools were popular for respondents, hence the relatively large number of suggestions for public officials to consider (questions 3 and 4, shown in Table 3). The only option ranked considerably lower than the rest was waiting to take action until additional research is done, which was not a popular choice. Additional work could help to identify the reasoning for this, and to better understand the opportunities for informed, strategic decisions that can be made at a later time (under a real options or another decision framework). How and where to incorporate sea level rise planning into other processes (whether in a stand-alone plan, incorporated into other plans, or through some other means) is likely to be a very local decision, given the wide array of potential items to include.

For protection priorities (finding 3), drinking water and electric power were both ranked with a median and mode of 5 (exceptionally high priority), making utilities key candidates for adaptation measures (questions 7 and 8, shown in Table 4). A number of other services (such as those related to transportation) ranked highly. However, it is also possible that officials will have a difficult time prioritizing certain areas and services over others, as most items polled were identified as high priorities, and many services are likely to be highly interdependent on each other.

For funding priorities (finding 4), the preferred method to identify funding was to hold public meetings (questions 10 and 11, shown in Table 5). Although public meetings may indeed be useful, this finding also poses the challenge that the identified funding sources (for example, using federal and state funds, which also ranked highly) may not be available when and in the quantities desired to meet the needs identified by public engagement. Local methods to raise funding (e.g., taxes) were potentially controversial and may pose challenges in gaining support, despite the greater level of local control.

For conflict resolution (finding 5), both discussions with preparedness experts and scientists were amongst the most popular choices (questions 13 and 14, shown in Table 6). A number of other means to prevent or resolve conflict (such as starting with measures that have the greatest public support, holding public meetings, and others) had similar levels of popularity, meaning that a number of conflict resolution methods may be acceptable to the public in sea level rise planning.

Finally, for adaptation responses (finding 6), most of the surveyed adaptation measures were generally considered acceptable by the respondents (questions 15 and 16, shown in Table 7). This included several measures that were expected to be controversial, such as preventing new development and removing existing development from vulnerable areas over time, were also generally acceptable. The only clearly unacceptable response of those studied was not providing assistance for areas at highest risk.

Impact of demographics

As mentioned in the results section, many demographics, such as income, education, and ethnicity had little impact on the distribution of responses for most components of most questions. Rather, the perceived level of environmentalism, preferred (public/private/mixture) funding sources, gender, and age impacted the most components of the questions. Although involving a diverse group of stakeholders across all demographics is essential to full engagement of a community, assuring a solid mixture of individuals affiliated with those demographics that have the greatest influence could be especially important in local sea level rise planning. By assuring, for example, that groups with differing viewpoints on environmental matters are represented in the process, there is greater potential for building buy-in through the process rather than ending in conflict.