Understanding Uncertainty in the Residential Landscape Using a Pilot Tree Risk Perception Scale

  • Arboriculture & Urban Forestry (AUF)
  • June 2026,
  • jauf.2026.015;
  • DOI: https://doi.org/10.48044/jauf.2026.015

Abstract

Background Residential damage from trees results in millions of dollars in insurance claims each year. In turn, insurance carriers have been increasingly forcing homeowners to prune or remove their yard trees for risk mitigation. Reduction of risk may or may not occur; however, the removal of tree canopy reduces the ecosystem benefits provided by trees.

Methods Our research employed a homeowner survey in the Southeastern United States to understand and describe influences on homeowners’ tree-related risk perceptions and subsequent decisions regarding their trees using a pilot tree risk perception scale.

Results Most respondents and their neighbors were influenced by their insurance carrier to prune or remove trees. Higher levels of tree-related risk perception significantly correlated with trust and satisfaction in their insurance. Respondents who had local insurance agents and personal relationships with their agents were significantly more concerned about tree risk than others. Tree-related risk perception negatively correlated with average tree canopy cover in the participants’ zip code.

Conclusions The study informs arborists and insurance carriers of the factors affecting homeowners’ tree-related risk perceptions and will therefore aid in policies and practices regarding mitigation and communication of the risks involved with privately owned trees. In the context of climate change, improved communication between these industries—both of which impact the community’s tree canopy—will become increasingly necessary to mitigate natural hazard-related risks to human health and safety while maintaining ecosystem services.

Keywords

Introduction

Increasingly extreme weather—including temperature fluctuations, intense storms, and wildfires—has surged insured losses, forcing the global insurance industry into a structural shift (Hoeppe 2016; Swiss Re Institute 2021). Many carriers have entered a “hard market” characterized by rising premiums and stringent underwriting, often withdrawing from high-risk regions or—in the context of the research presented here— increasing scrutiny of yard trees (Auer 2021; Wagner 2022; Hofmann and Sattarhoff 2023; Apodaca 2024; Cox 2024; Darmiento 2024; Honeycutt 2024; Thomasson 2025). In response, states have intervened through wind pools or reinsurance mandates (Crowley 2023; California Department of Insurance 2024), while initiatives like Britain’s Flood Re provide international parallels (Donnellan 2025). This escalating influence on residential landscapes is becoming a global issue with far-reaching environmental implications (Torpy 2024; Rawlinson 2025; Recamara 2025).

As climate change shifts hazard patterns, insurers are incorporating new technologies, from drones and satellite imagery to forward-looking, AI-driven catastrophe models (Klein 2009; Frazier 2021; Gray 2021; Wagner 2022; Gupta and Venkataraman 2024). In the case of residential trees, insurer assessments often overlook arboricultural standards (Koeser and Smiley 2017; Klein et al. 2019; TCIA 2023). While tree risk science is based on condition, biomechanics, and experts’ experience, insurers’ lack of specialized training may lead to recommendations that unnecessarily degrade the urban canopy. The effect on homeowners’ risk perceptions—the subjective judgment of a hazard’s severity—remains unclear (Darrow 2024; Eaglesham 2024).

Unlike technical assessments, lay perceptions are intuitive and driven by trust, effect, and experience (van der Linden 2015). Factors such as perceived anthropogenic cause or controllability shape concern, which must be personalized to motivate action (Spence et al. 2012). Personal experience can heighten perceived probability (Kousky 2017) or normalize risk through repeated low-severity exposure (Botzen et al. 2015). Furthermore, the balance between perceived control and efficacy determines whether individuals engage in mitigation or maladaptive denial (Kowalski et al. 2023). Psychological discounting of continuous risks (Calvani et al. 2025), amenity tradeoffs (Gordon et al. 2010), and social amplification through neighbors or media (Bonfanti et al. 2024) further complicate these decisions.

This research addresses two objectives: (1) documenting insurance influences on tree-related decisions, and (2) identifying factors shaping homeowners’ risk perceptions. This work is critical, as insurers may be indirectly degrading the canopy and amplifying climate effects. While tree failure may seem mundane compared to disasters such as catastrophic wildfire, its daily impact warrants scholarly attention. Ultimately, findings point to opportunities for improved communication between the tree care and insurance industries to encourage sustainable management of urban forest systems as well as accomplish economic objectives.

Literature Review

Risk Perception

Risk is not a direct representation of physical hazard but a construct developed to interpret and manage uncertainty (Slovic 1992). Traditional natural hazard management relied on a technocratic or rational actor model, which assumed individuals would rationally align behavior with expert assessments of probability and consequence when provided with accurate information. Catastrophic wildfires in California (2017–2018), Australia (2019–2020), and Southern Europe (2021–2023) have exposed the limitations of this information-deficit paradigm, shifting attention toward the socio-psychological foundations of risk perception, particularly how individuals evaluate probability, experience concern, and perceive control (Wachinger et al. 2013).

To understand why the technocratic model often fails, it is necessary to examine how individuals move beyond raw data to form their own internal versions of threat. Contemporary literature conceptualizes risk perception as a socially constructed process rather than a direct interpretation of objective hazard probability. Perceived probability of harm is filtered through demographic characteristics, political ideology, age, and education (Gustafsod 1998; Lujala et al. 2015; Bronfman et al. 2020; Rosi et al. 2021). Yet, this body of work demonstrates that perceived probability alone is a weak predictor of protective behavior unless accompanied by emotional engagement and perceived control. Psychometric and cultural approaches further show that trust in institutions—public and private— influences whether individuals accept expert probability estimates or disengage entirely, viewing disaster outcomes as inevitable (Xue et al. 2014; Gauldin et al. 2025b).

Because these social filters transform statistics into personal meaning, the resulting emotional response becomes the primary engine for action. A growing literature demonstrates that concern—often expressed as worry, fear, or anxiety—is a central driver of risk perception and action. Hazards perceived as anthropogenic elicit heightened concern and anger compared to purely natural events, shaping both policy preferences and individual mitigation behavior (Carroll et al. 2004; Picou et al. 2004). Some events, such as wildfire, occupy a hybrid space compared with earthquakes and tornados, which are universally perceived as uncontrollable acts of God (Fox-Glassman and Weber 2016). Lidskog et al. (2019) emphasize that wildfires are often viewed as managed events because humans can start fires and (theoretically) put them out. After flooding, public demand often focuses on infrastructure (levees); after a fire, demand focuses on accountability and punishment for ignition sources, often diverting attention from personal preparedness (Zia et al. 2023).

However, this emotional engagement is a delicate balance. While concern can drive accountability, its impact depends heavily on whether the threat feels immediate or abstract. For example, increasing attribution of wildfire severity to climate change has elevated general concern; however, it can also reduce personal concern if individuals perceive the problem as global rather than local, leading to passivity (Skagerlund et al. 2020). These findings suggest that concern must be localized and personalized to translate into household-level action, particularly in contexts such as tree risk where damage may be episodic but consequential (Peters et al. 2006; Spence et al. 2012). Recently, successful risk communication has shifted from presenting data tables to using narrative and visual imagery that triggers appropriate levels of concern without inducing paralysis (Dash and Gladwin 2007).

While communication strategies aim to spark this localized concern, an individual’s own history with disaster often acts as a more powerful, albeit inconsistent, teacher. Direct experience with wildfire shapes perceived probability in complex and sometimes counter-intuitive ways. Severe experiences involving loss or fear can heighten perceived probability and motivate mitigation such as insurance purchase (McGee et al. 2009; Gan et al. 2014), whereas repeated exposure to low-severity events can normalize risk, reducing concern and diminishing perceived probability over time (McCaffrey 2004; Champ and Brenkert- Smith 2016). Unlike catastrophic floods, which can permanently elevate perceived probability and insurance uptake (Kousky 2017), chronic wildfire exposure may lead homeowners to incorporate risk into daily life, weakening concern and preparedness. Past experiences may also distort probability judgments under changing climatic conditions, resulting in systematic underestimation of current risk (Ryan 2010; Botzen et al. 2015; Bakkensen and Barrage 2022).

Given the complex interplay between internal emotions and external experiences, researchers have sought unified models to explain how these variables coalesce into a single decision. One approach, the Climate Change Risk Perception Model (CCRPM) (van der Linden 2015), provides a unified framework that directly aligns with these constructs by integrating cognitive evaluations of probability, experiential components of concern, and sociocultural influences on perceived control. Tested on a nationally representative UK sample, the model explained nearly 70% of the variance in perceived risk. Experiential factors— particularly effect and personal experience—were stronger predictors of concern than knowledge-based assessments of probability. Importantly, the model distinguished between societal risk and personal risk, demonstrating that high perceived probability at the societal level does not necessarily produce personal concern or motivate individual action without a sense of control.

This sense of control identified in the CCRPM is often the final pivot point between a homeowner taking action or falling into a state of denial. Perceived control is a defining feature of wildfire risk perception, in particular. Homeowners in the wildland–urban interface often exhibit optimism bias, believing they are less vulnerable than others, while simultaneously expressing high internal control—the belief that actions such as defensible space or tree management can meaningfully reduce risk (Martin et al. 2009). However, when perceived probability and concern are high but perceived control is low, individuals frequently engage in maladaptive coping, including denial and avoidance (Bubeck et al. 2012). Conversely, overestimation of control can lead to complacency and under-preparedness (Martin et al. 2009). These dynamics underscore the importance of measuring both internal control (self-efficacy) and external control (feelings of helplessness).

The unique physical nature of fire further complicates this sense of control, often allowing for psychological loopholes that do not exist in other natural disasters. Wildfires differ from other hazards in ways that complicate perceptions of probability and control. Unlike floods, which are geographically bounded and mapped, wildfire risk is continuous and probabilistic, allowing homeowners greater psychological latitude to discount probability and exaggerate control (Botzen et al. 2019). Difficulty conceptualizing ember transport further weakens homeowner probability assessments, believing that if they are not touching the forest, they are safe (Calvani et al. 2025). When wildfire risk is perceived as voluntary, individuals downplay both probability and consequence to justify amenity-based living, reinforcing an amenity–risk trade-off that suppresses concern and mitigation behavior (Jansen et al. 2017). Many of these characteristics can apply to residential tree spaces, such as distance to house, wind-throw effect, and the voluntary nature of tree maintenance.

Even these individual psychological calculations do not happen in a vacuum; they are constantly amplified or dampened by the surrounding community. Risk perception is further shaped by social context. Observing neighbors engage in mitigation increases perceived probability and concern through risk interdependency while also reinforcing beliefs about the effectiveness of individual control (Brenkert-Smith et al. 2013). Media coverage often amplifies immediate losses, elevating concern, while attenuating longer-term ecological processes that inform realistic probability assessments (Kasperson et al. 2022). Likewise, this can be observed in media interpretations of tree failure after major storms with dramatic images of trees interrupting transportation or damaging property. In turn, trust in institutions (public and private) moderates all 3 constructs: probability estimates, concern, and perceived control. Communities that perceive shared values with institutions are more likely to accept risk assessments and engage in mitigation, whereas distrust leads to rejection and disengagement (Earle 2010; Siegrist 2021; Bonfanti et al. 2024).

Ultimately, the synthesis of these social and psychological factors points toward a more nuanced approach to evaluating how people live with environmental danger. Taken together, the risk perception literature consistently emphasizes 3 interrelated constructs: perceived probability of harm, degree of concern or worry, and beliefs about personal and external control. These constructs recur across wildfire, climate change, flooding, and other natural hazards, directly shaping household-level decision-making. Accordingly, understanding how homeowners evaluate tree-related risk requires measurement tools that explicitly capture probability, concern, and control within the residential context.

Risk Perception Scale

The wildfire literature influenced our approach, because, like tree failure, wildfire is often considered a “natural” hazard, and the risk perception of wildfire body of work is extensive. While wildfire risk perception scales are not uncommon, public health scales such as diabetes risk perceptions have been much more commonly tested in research due to the prevalence of such diseases worldwide. In addition, diabetes, like tree risk, is often considered a chronic condition. Consistent with the wildfire and public health literatures, homeowner risk perception in this study is conceptualized as a multidimensional construct composed of perceived probability, degree of concern, and perceived control. This framing reflects established research demonstrating that responses to hazards are shaped less by objective risk than by how individuals interpret the likelihood of harm, emotionally respond to that possibility, and assess their capacity to influence outcomes.

Within the wildfire literature, probability is commonly operationalized as homeowners’ perceived likelihood that a hazard will affect their property, often framed as damage resulting from extreme weather or fire events (Blanchard and Ryan 2007; Brenkert-Smith et al. 2013; Fischer et al. 2014; Meldrum et al. 2015). Hall et al. (2022) integrated these approaches by measuring both omnibus risk to the home and perceived likelihood of wildfire damage. This body of work consistently shows that perceived probability alone rarely motivates protective action.

Concern captures the affective dimension of risk perception, including worry or anxiety about potential harm. In wildfire research, concern reflects the extent to which homeowners are troubled by nearby hazards (Hall et al. 2022), while public health studies similarly emphasize concern as a mechanism through which probabilistic threats become personally salient (Rochefort et al. 2020). Together, these findings indicate that concern mediates the relationship between perceived probability and behavioral response.

Perceived control moderates how probability and concern translate into action. Wildfire studies frequently reference control through beliefs about preparedness and mitigation, while public health research distinguishes between internal control (belief in one’s ability to reduce risk) and external control (perceptions of helplessness)(Rochefort et al. 2020). Across both literatures, low perceived control amplifies fatalism, whereas excessive internal control may suppress concern through optimism bias.

Drawing directly from these constructs, we developed and piloted the Homeowner’s Tree Risk Perception Scale (HTRPS) to measure tree-related risk through probability, concern, and control. Probability items assess perceived likelihood of tree-related damage; concern items capture emotional responses to that likelihood; and control items distinguish internal and external beliefs about managing tree risk. By aligning these constructs within a residential tree context, the HTRPS provides a theoretically grounded tool for examining how risk perceptions shape tree management decisions and, ultimately, residential canopy outcomes.

Research Hypotheses

We tested the following hypotheses relative to the study objectives described previously:

Objective 1—Insurance carrier influences:

(H1) Insurance carriers influence respondents’ decisions to prune or remove trees.

(H2) Respondents more often respond to insurance company requests by pruning/removing their trees than challenging the company.

Objective 2—Factors that influence homeowners’ risk perceptions:

(H3) Risk perception scores correlate with distance from the coast.

(H4) Those who are more concerned about extreme weather are more concerned about the risk their trees pose to their property than those who are less concerned about extreme weather.

(H5) Those who live in areas where there is more tree canopy will have higher tree-related risk perception due to greater interaction with trees than those who live in areas with less canopy.

(H6) Homeowners who report higher trust and satisfaction with their insurance carrier will report higher levels of perceived tree-related risk.

Methods

Survey Distribution and Design

A survey was conducted in the Southeastern region of the United States (Figure 1)(UGA IRB ID: PROJECT 00010111). We selected this area of the country not only due to budgetary constraints but also due to insurance carriers developing new restrictions or withdrawing from markets as a result of increasing natural hazard risk, primarily tornados and hurricanes (NOAA 2024). The survey consisted of 31 questions (Appendix), including 10 demographic questions, although logic behavior allowed participants to skip some of the substantive questions. Substantive questions focused on the following question blocks: (1) the participant’s concerns and actions regarding their yard trees; (2) if homeowners insurance carriers influenced or would influence their yard tree management; and (3) communication with their homeowners insurance carrier. A pilot survey was conducted to test wording and structure prior to survey distribution.

Study area.
Figure 1.

Study area.

An initial question asked about ownership of yard trees at the time of the survey or any time prior to the survey with a negative response triggering skip logic to the questions about communication with insurers. For participants who responded affirmatively, follow-up questions asked if their insurance had influenced them to prune and/or remove trees and if they knew of other people who had been influenced. Also, participants were asked if they had a local agent (compared with a remote service) that serviced their insurance and if they had a personal relationship with that agent. Respondents answered questions about the limiting factors to their own tree maintenance (e.g., time, money, lack of knowledge, or restrictive ordinances) and the approximate distance of the trees from their house. One question addressed climate change based on their level of concern about increasingly frequent severe weather. The survey ended with demographic and socioeconomic questions about their income, length of residency, household size, political opinions, gender, race, age, participation in wind pools, and postal code.

Homeowner’s Tree Risk Perception Scale

We adapted a risk perception scale from previously published scales in the wildfire and public health literatures (Boateng et al. 2018; Rochefort et al. 2020; DeVellis and Thorpe 2021; Hall et al. 2022). We began by creating a list of 21 initial questions aimed at measuring the constructs of omnibus risk, probability, consequence, concern, worry, external control, internal control, and optimistic bias with 2 to 4 questions for each construct (Rochefort et al. 2020; Hall et al. 2022). From this list, we selected the most appropriate questions for this study, resulting in 8 items. Each of these questions used a 5-point Likert scale to measure the construct.

We used Cronbach’s alpha in RStudio (Posit 2024) to measure internal consistency and the usefulness of each question (Taber 2018). The initial alpha was calculated at 0.7 but reported a warning that one question (Do you think that you can effectively take actions to reduce the risks of a tree or tree limb falling on your house?) was negatively correlated with the first principal component and likely needed to be reversed. The question had been scaled opposite (5-1) to the other scale questions because an increase of internal control, or the ability to prepare for and reduce tree-related risk of harm to the participant’s house, was expected to lower the overall risk perception of the participant. Reversing the response order, as RStudio suggested, would not be conceptually consistent, as increasing internal control, or self-efficacy, was associated with decreased levels of risk perception. To increase the internal validity of the scale, while not sacrificing conceptual consistency, we removed the question. Further, we justified this decision based on the 26 wildfire scales referenced by Hall et al. (2022) also not measuring internal control. The final Cronbach’s alpha of 0.85 ensured adequate scale validity (Tavakol and Dennick 2011). We then summed each of the scale responses to create a cumulative risk perception score for each of the respondents. After the removal of the question, the maximum score possible for the scale was 28 and the minimum score possible was 0.

Data Collection

The survey was conducted during September and October of 2024, during hurricane season in the Southern United States, which runs from June to November (National Hurricane Center and Central Pacific Hurricane Center 2024). We distributed the survey through Mechanical Turk (MTurk), an online crowdsourcing platform (www.mturk.com). MTurk is operated by Amazon Web Services (Seattle, WA, USA) and has been demonstrated as a viable option for behavioral and psychological research (Paolacci et al. 2010; Keith et al. 2017; Thomas and Clifford 2017).

A pilot survey (n = 57) explored any irregularities or response issues related to question phrasing or order. After checking these pilot results and making any necessary revisions, we distributed the survey out to the rest of the respondents. A total of 1,800 participants (i.e., workers) were recruited. Workers completed the survey on Qualtrics (2025) after the task was posted to the MTurk platform. Workers completed the survey in approximately 6 min, for which they were paid $1.50 USD, a typical rate for MTurk. As noted above, screening questions addressed ownership of a house and homeowners insurance. A house was defined as a primary or secondary residence consisting of an individual dwelling that was not an apartment, condominium, or townhouse. A negative response to either of these questions concluded the survey.

The final number of completed surveys in the study area was 500. Of respondents, 41 had either never had trees in their yard, no longer had trees in their yard, or did not finish answering the scale questions and therefore could not be used in the tree risk perception scale analysis. This left 459 responses that could be used for analysis of the HTRPS (Table 1). The sample is representative of homeowners in the Southeastern USA; however, potential biases related to MTurk workers should be considered. As such, results should be interpreted with caution and may not be generalizable to a wider population.

View this table:
Table 1.

Completed surveys.

Data Analysis

Survey responses were linked to an ArcGIS Pro (Esri, Redlands, CA, USA) file. Respondents’ postal codes were then linked to environmental data using the Geocode function in ArcGIS Pro which places a point at the centroid of the postal code polygon (ESRI 2025a). We geocoded the data so we could examine the relationship between risk perception and nearby canopy cover along with participants’ distance to the coast. We employed tree canopy cover raster data from the National Land Cover Dataset (NLCD) which is based on satellite imagery to determine the percentage of tree canopy cover for 30-m by 30-m pixels for the USA (Dewitz 2023). We then approximated canopy percentage in the areas near each respondent’s self-reported postal code after layering polygons for every US postal code (ESRI 2025b). We then created a North American coastline feature class (ESRI 2025b). We made a Euclidean distance raster representing distance from the coast and extracted the values to the data points. We linked the geodatabase file to RStudio for statistical analysis (Posit 2024).

Significance for all tests was α ≤ 0.05. After performing a Shapiro-Wilk test of normality on the risk perception summative scale (the dependent variable), we found the cumulative risk scores to be not normally distributed with a positive skew (skew = 0.464, W = 0.970, P ≤ 0.000). Because of this, we used nonparametric tests to compare the tree-related risk perception scores among various categories within the data.

Predictive Model

We employed logistic regression to explore the relative importance of respondents’ socioeconomic status, their spatial location and relative tree canopy near their property, influences of their insurance, their communication with insurance, and their concern for increasingly severe weather to explain their HTRPS scores. The composite variable was dichotomized based on the theoretical midpoint of the scale (14.0). This cut-point was selected to preserve the semantic meaning of the response anchors, distinguishing between participants who, on average, affirmed the construct (Mean > 14) versus those who rejected or remained neutral toward it (Mean ≤ 14)(MacCallum et al. 2002).

Data satisfied assumptions of independence of observations as the dependent and independent variable categories were mutually exclusive. Additionally, we determined that no variables in the models were multicollinear with other variables as the Variance Inflation Factor (VIF) for the variables in the model were within adequate ranges (VIF for all variables < 1.6). We did not use replacement imputation for missing values for the successive stages of the model. We successively added the indicators to the model to observe changes and improvements to the model (i.e., change of coefficients and model fit). We created five different models starting first with a base model that used only socioeconomic and demographic variables (Model 1). The next model added in the spatially related variables (Model 2). The next model added variables related to insurance carriers’ influence on the participants yard tree management (Model 3). The next model added variables related to the participants’ communication with their insurance company (Model 4). The final model added a variable about concern for increasingly frequent severe weather (Model 5) (Table 2).

View this table:
Table 2.

Logistic regression of 5 models predicting higher HTRPS (Homeowner’s Tree Risk Perception Scale): (I) socioeconomic background questions only; (II) addition of spatial variables; (III) insurance influence questions; (IV) communication with insurance questions; (V) addition of severe weather concern question.

Results

Descriptive Statistics

On average, the respondents were 35 years old, male (63.8%), and white (94.7%)(Appendix). Of the 500 participants, the top represented states included Florida (43.2%), Georgia (13.8%), and North Carolina (11.2%). The other states individually represented less than 10% of the total respondents. Respondents’ locations averaged 75.89 miles from the coast, and average tree canopy was 36.23% of their postal code area (Figure 2). The average reported household size had 1 to 4 members (77.8%), and the average reported length of residence was 16 to 20 years (< 5 years = 7.2%; 6 to 10 years = 24.4%; 11 to 15 years = 16.3%; 16 to 20 years = 11.2%; 21 to 25 years = 10.6%; 26 to 30 years = 13.1%; 30 < years = 17.3%). Most respondents claimed to have either a moderate (30.1%) or conservative (26.9%) political view. Those that reported having liberal or very liberal political views represented 20.1% of the respondents. Most of the respondents (73.2%) reported having trees closer than about 40 ft (12.2 m) from their home. Time and money were the key limiting factors for determining tree maintenance (34.4% and 33.3%, respectively).

Map of the Canopy Zonal Analysis.
Figure 2.

Map of the Canopy Zonal Analysis.

Most respondents’ (84.11%) decisions to prune or remove trees on their property had been influenced by their insurance carrier, and many respondents knew others who had hired a tree company in reaction to the insurance requests (52.99%). About one-third of respondents (29.30%) changed insurance companies because of the tree pruning/removal requests. The majority reported using a local agent (90.65%), with 78.66% of them reporting having a personal relationship with their insurance agent. The average reported level of trust and satisfaction with their insurance company by the respondents was 2.35 and 2.37 respectively, and their concern for increasingly severe weather was 2.39 (Likert scale for each: 0 to 4).

Risk Perception Differences and Correlations

Respondents having a local agent had a significantly higher HTRPS than those without one (median score of 16 vs. 14.5 [P ≤ 0.009]). There was a statistical difference between the two groups despite the closeness in scores, suggesting those with a local agent were more risk averse than their counterparts. Additionally, those who reported having a personal relationship with their agent had a significantly higher HTRPS than those who did not have a personal relationship with their agent (median score of 17 vs. 15 [P ≤ 0.0000]).

Those who reported being influenced by their insurance carrier to prune or remove trees had significantly different HTRPS than those who reported not ever being influenced by their insurance in the past (P ≤ 0.0000). Additionally, those who reported being influenced 3 or more times by their insurance carrier had significantly higher HTRPS than those who reported being influenced by their insurance less (1 to 3 times vs. 3 or more, P ≤ 0.0000; none vs. 3 or more, P ≤ 0.0000). Study participants who reported changing insurance companies had significantly lower tree-related HTRPS than the other groups (changed insurance vs. removed/pruned trees, P ≤ 0.0000; considering removing/pruning, P ≤ 0.0063; refused to remove/prune, P ≤ 0.0000). There were significantly different HTRPS scores among respondents with different political ideologies with Very Liberal and Very Conservative having the highest scores (median scores of 19 and 20, respectively) and Moderate with the lowest (median score of 15)(Very Conservative vs. Conservative, P ≤ 0.000; Very Conservative vs. Liberal, P ≤ 0.004; Very Conservative vs. Moderate, P ≤ 0.000; Very Liberal vs. Moderate, P ≤ 0.025).

There was no significant correlation between distance from the coast and HTRPS (P ≤ 0.508). There was a weak significant negative correlation (rho = –0.15) between average tree canopy and HTRPS (P ≤ 0.0013), meaning that as average tree canopy increased, HTRPS generally decreased. There was a significant moderate positive correlation between those with higher reported concern for increasingly frequent and severe weather and HTRPS (rho = 0.57, P ≤ 0.0000). This means as their reported concern for severe weather increased, so did their HTRPS. Similarly, trust and satisfaction with their insurance company demonstrated a significant moderately positive correlation with participants’ HTRPS (rho = 0.492, P ≤ 0.0000 and rho = 0.465, P ≤ 0.0000, respectively). As participants reported higher levels of trust and satisfaction with their insurance companies, they also reported higher HTRPS.

Predicting Homeowners’ Tree Risk Perceptions

We added predictive variables to the logistic regression model for a total of 5 models (Table 2). The largest improvement in model fit indicators happened between models II and III with an increase in McFadden’s R2of 0.489 and a decrease in Residual Deviance of 276.46. In the first model, age, household size, and the distance of trees to homeowners’ houses were significant variables in predicting HTRPS. Higher HTRPS were more likely with higher ages, greater distances of trees to the homeowner’s houses (P < 0.001), and smaller household sizes. This model explained approximately 8.6% of the variation in tree risk perception.

Model 2 showed little improvement as compared to Model 1 when adding spatial components of proximity to the coast, average tree canopy, and state location. As with the previous model, age, household size, and the distance of trees to homeowners’ properties were the variables that significantly explained some of the variation in HTRPS. Odds ratios were consistent with the previous model. Model 3 introduced the variables related to the insurance companies’ influence on the respondents’ decisions and management of their trees. This model explained 57.7% of the variability in homeowners’ tree risk perceptions. Age, household size, and distances of trees to the homeowners’ houses stayed significant contributors to the model, with the addition of trust in insurance to make decisions relating to their yard tree maintenance becoming the most significant predictor for this model. Odds ratios of the first 3 variables stayed consistent with the previous models. Notably, higher HTRPS was more likely with greater levels of trust in their insurance company to make decisions regarding their yard tree management (P < 0.001).

Model 4 increased the ability of the model to predict 65.4% of the variability in HTRPS. This model added two more variables: one about satisfaction with their insurance company, and the other about having a personal relationship with their insurance agent. The significant predictors for this model were age (P < 0.01), trust in their insurance (P < 0.001), and satisfaction with their insurance company (P < 0.001). Higher HTRPS for homeowners was more likely with higher age and greater levels of trust and satisfaction in their insurance company.

Model 5, the final model, was the only significant model and increased the predictability of the model to 69.0% of the variability in HTRPS. This model added a final question about concern for increasingly frequent severe weather. The significant predictors for this model were age (P <0.01), trust (P <0.01), satisfaction (P <0.01), and severe weather concern (P <0.001). Higher HTRPS for homeowners was more likely with higher age, greater levels of trust, satisfaction in their insurance company, and greater concern for increasingly frequent severe weather.

Discussion

Contrary to much of the risk perception literature, this study found no consistent differences in tree-related risk perception across age and gender once insurance-related variables were introduced into the models (Table 2). Prior research has frequently documented demographic variation in perceived risk across domains such as driving, smoking, and public health hazards (Viscusi 1991; Flynn et al. 1994; Rhodes and Pivik 2011; Rosi et al. 2021). The absence of such effects here suggests that tree-related property risk may operate within a distinct decision domain, where institutional cues—particularly from insurance carriers— override demographic tendencies. This interpretation aligns with Bonem et al. (2015), who argue that risk preferences vary by motivation and context rather than reflecting stable individual traits. Sampling limitations, including the skew toward male respondents and middle-aged homeowners, may also partially explain these results; however, the dominance of insurance-related predictors across models suggests a substantively meaningful pattern rather than a statistical artifact.

Although not central to the original hypotheses, political ideology emerged as a significant correlate of the HTRPS, with respondents identifying as “Very Liberal” or “Very Conservative” reporting the highest perceived risk and moderates reporting the lowest. This nonlinear pattern mirrors findings in climate risk perception research, where ideological extremes often express heightened concern through different causal narratives (Lee et al. 2015; van der Linden 2017). This result reinforces sociocultural theories of risk perception, which emphasize that concern is shaped by values and identity as much as by perceived probability. However, the relatively small proportion of liberal respondents warrants caution in interpretation.

A central objective of this research was to examine how insurance companies influence homeowner decision-making and, by extension, the urban tree canopy. Findings addressing Hypotheses 1 and 2 demonstrate that insurance carriers exert substantial influence over tree management decisions, with most respondents complying with pruning or removal requests rather than challenging them or seeking alternative assessments. This pattern reflects broader structural shifts in the insurance industry, including the transition to a hard market, increased underwriting scrutiny, and reliance on remote risk assessments (Hoeppe 2016; Wagner 2022; Hofmann and Sattarhoff 2023). While urban trees represent a relatively small proportion of total forest cover, cumulative loss on private residential land can significantly reduce ecosystem services critical to urban health and climate adaptation (Frazier 2021; Gray 2021).

Tree canopy cover exhibited a weak but significant negative relationship with HTRPS when examined independently, suggesting that homeowners in more heavily treed areas reported lower perceived risk. This finding contrasts with studies showing that direct experience with hazards increases perceived probability and concern (Wachinger et al. 2013; Lujala et al. 2015) but is consistent with research on risk normalization and desensitization following repeated low-severity exposure (McCaffrey 2004; Champ and Brenkert-Smith 2016). Similar to Mayer et al. (2017), greater environmental exposure alone did not translate into heightened risk perception, indicating that experience influences probability judgments only when coupled with fear or loss (McGee et al. 2009). This suggests that tree-related risk may function as a chronic, background hazard—analogous to seasonal wildfire smoke—rather than as a salient, acute threat.

Distance-based variables further complicate this interpretation. Proximity to coastal hazards did not significantly predict tree risk perception, despite extensive literature linking physical and psychological proximity to heightened concern (Spence et al. 2012; Balžekienė et al. 2024). This may reflect a cognitive decoupling between storm exposure and tree failure, underscoring the challenge homeowners face in translating complex, multihazard environments into coherent probability assessments. Similarly, greater distance between trees and homes was associated with higher HTRPS, contrary to expectations. One plausible explanation is selection bias: homeowners with high concern may have already removed closer trees, leaving a residual population that perceives remaining trees—often larger or less manageable— as more threatening. Alternatively, homeowners with trees close to structures may accept the risk due to amenity value or economic constraints, reflecting an amenity-risk trade-off documented in wildfire and flood contexts (Meldrum et al. 2015).

Trust emerged as one of the most influential predictors of tree-related risk perception. Respondents reporting higher trust and satisfaction with their insurance carriers also reported higher HTRPS scores, consistent with literature demonstrating that trust amplifies acceptance of expert risk framing (Earle 2010; Covello 2021; Siegrist 2021). Having a personal relationship with a local agent further elevated perceived risk, suggesting that relational trust—not merely institutional credibility—plays a central role in shaping concern and perceived probability. This relationship may indicate the homeowner was more likely to address the risk in a positive way instead of fearing a policy change. However, this dynamic may also reflect authority bias, whereby trusted institutions implicitly legitimize hazard salience. Importantly, this finding does not imply that trust leads to maladaptive fear; rather, it suggests that trusted insurers may successfully activate concern without triggering denial, thereby increasing compliance with mitigation requests.

Respondents who had pruned or removed trees at their insurer’s request exhibited the highest HTRPS scores, aligning with established links between high risk perception and protective action (Peacock 2003; Lindell and Hwang 2008). From a behavioral perspective, insurers occupy a powerful position within the risk communication landscape: by shaping perceived probability and concern through narratives, imagery, or third-party validation (e.g., arborists), they can increase compliance while framing decisions as voluntary safety measures rather than mandates. However, widespread amplification of tree-related risk could also accelerate canopy loss, reinforcing a structural tension between insurer objectives (risk minimization) and municipal goals (shade, stormwater control, and climate mitigation).

This tension highlights the importance of perceived control. Risk perception theory consistently shows that high concern without perceived control leads to fatalism and avoidance (Bubeck et al. 2012). In contrast, calibrated self-efficacy promotes adaptive behavior. Collaborative approaches that involve certified arborists, transparent standards, and shared values may help recalibrate control perceptions, preventing insurers from becoming de facto drivers of unnecessary tree removal (Koeser and Smiley 2017; TCIA 2023).

Finally, these findings point toward applied solutions. Behavioral tools such as simplified insurance products or default coverage options may help reduce cognitive load while supporting risk transfer rather than risk elimination (Robinson and Botzen 2019). In addition, existing programs such as Firewise USA and Fortified Homes demonstrate how standardized frameworks can align homeowner behavior, insurance incentives, and hazard mitigation (IBHS 2024; NFPA 2024). A comparable initiative—tentatively described as Fortified Landscapes—could integrate arboricultural science, insurance underwriting, and residential design to manage tree risk without sacrificing canopy benefits (Gauldin et al. 2025a). Such programs would not deny that trees pose risks but would contextualize those risks within statistical reality, emphasizing trees’ protective functions against wind and water while addressing legitimate structural concerns.

Conclusion

This study reinforces that tree-related risk perception is not driven primarily by demographics or exposure but by institutional trust, affective concern, and perceived control. For example, while age emerged as a statistically significant predictor in the final logistic regression model, its practical impact was negligible, with coefficients below 0.1 across all models. Similarly, no significant differences were observed by gender, and risk perception scores did not correlate with coastal proximity, leading to the rejection of Hypothesis 3.

However, the results supported Hypotheses 1 and 2, revealing that insurance carriers are a dominant driver of canopy management. A substantial 84.11% of respondents reported their tree maintenance decisions were influenced by their insurance company. Rather than challenging these mandates, 50.88% of participants had either performed the requested pruning/removal or were actively considering it. Risk perception was also closely tied to broader environmental anxieties; supporting Hypothesis 4, respondents with higher concern for severe weather reported significantly higher HTRPS scores.

Contrary to Hypothesis 5, a weak negative correlation was found between tree canopy density and HTRPS, suggesting that increased exposure to trees may normalize risk perception. This was further evidenced by the finding that homeowners were more likely to report higher HTRPS as the distance between their house and the nearest trees increased. Finally, the study confirmed that trust and satisfaction with insurance carriers are significant predictors of risk perception, suggesting that homeowners rely heavily on carrier-provided risk assessments even if these are not supported by arboriculture science.

The HTRPS provides a critical tool for understanding the drivers of homeowner tree management, a domain previously dominated by studies of professional perceptions. Because property owners serve as the primary deciders of urban canopy maintenance, understanding their subjective risk evaluations is essential for both risk reduction and ecological conservation. Future research should apply the HTRPS across diverse geographic contexts and among varied stakeholders to further refine strategies for resilient urban forest management. Ultimately, the research demonstrates that insurance companies now play a central role in shaping these constructs. Whether that influence degrades or sustains the urban forest will depend on how risk is communicated, by whom, and with what underlying values.

Conflicts of Interest

The authors reported no conflicts of interest.

Acknowledgements

This research was funded by the USDA National Institute of Food and Agriculture under Grant Proposal Number: 2023-05134. The views represented in this study do not represent the opinions of the grantor.

Appendix

View this table:
Table S1.

Survey questions*. HBRPS-4 (Homeowner Bushfire Risk Perception Scale); RPS-DD (Risk Perception Scale for Developing Diabetes).

View this table:
Table S2.

Descriptive statistics for variables used in the analysis of the drivers and outcomes of tree related risk perception.

Literature Cited

  1. ↵
    Apodaca T. 2024 October 22. Living in the City of Trees comes with a cost: Maintenance to maintain homeowners insurance. Sacramento (CA, USA): CBS News. https://www.cbsnews.com/sacramento/news/living-in-the-city-of-trees-comes-with-a-cost-maintenance-to-maintain-homeowners-insurance
    Google Scholar
  2. ↵
    Auer MR. 2021. Considering equity in wildfire protection. Sustainability Science. 16:2163–2169. https://doi.org/10.1007/s11625-021-01024-8
    Google ScholarOpenURLPubMed
  3. ↵
    Bakkensen LA., Barrage L. 2022. Going underwater? Flood risk belief heterogeneity and coastal home price dynamics. The Review of Financial Studies. 35(8):3666–3709. https://doi.org/10.1093/rfs/hhab122
    Google ScholarOpenURL
  4. ↵
    Balžekienė A, Echavarren JM., Telešienė A. 2024. The effect of proximity on risk perception: A systematic literature review. Current Sociology. 73(6):00113921241250047. https://doi.org/10.1177/00113921241250047
    Google ScholarOpenURL
  5. ↵
    Blanchard B., Ryan RL. 2007. Managing the wildland-urban interface in the Northeast: Perceptions of fire risk and hazard reduction strategies. Northern Journal of Applied Forestry. 24(3):203–208.
    Google ScholarOpenURLWeb of Science
  6. ↵
    Boateng GO., Neilands TB., Frongillo EA., Melgar-Quiñonez HR., Young SL. 2018. Best practices for developing and validating scales for health, social, and behavioral research: A primer. Frontiers in Public Health. 6:149. https://doi.org/10.3389/fpubh.2018.00149
    Google ScholarOpenURLPubMed
  7. ↵
    Bonem EM., Ellsworth PC., Gonzalez R. 2015. Age differences in risk: Perceptions, intentions and domains. Journal of Behavioral Decision Making. 28(4):317–330. https://doi.org/10.1002/bdm.1848
    Google ScholarOpenURL
  8. ↵
    Bonfanti RC., Oberti B., Ravazzoli E., Rinaldi A., Ruggieri S., Schimmenti A. 2024. The role of trust in disaster risk reduction: A critical review. International Journal of Environmental Research and Public Health. 21(1):29. https://doi.org/10.3390/ijerph21010029
    Google ScholarOpenURL
  9. ↵
    Botzen WJW., Deschenes O., Sanders M. 2019. The economic impacts of natural disasters: A review of models and empirical studies. Review of Environmental Economics and Policy. 13(2):167–188. https://doi.org/10.1093/reep/rez004
    Google ScholarOpenURLCrossRef
  10. ↵
    Botzen WJW., Kunreuther H., Michel-Kerjan E. 2015. Divergence between individual perceptions and objective indicators of tail risks: Evidence from floodplain residents in New York City. Judgment and Decision Making.. 10(4):365–385. https://doi.org/10.1017/S1930297500005179
    Google ScholarOpenURL
  11. ↵
    Brenkert-Smith H., Dickinson KL., Champ PA., Flores N. 2013. Social amplification of wildfire risk: The role of social interactions and information sources. Risk Analysis. 33(5): 800–817. https://doi.org/10.1111/j.1539-6924.2012.01917.x
    Google ScholarOpenURLPubMed
  12. ↵
    Bronfman NC, Cisternas PC, Repetto PB, Castañeda JV, Guic E. 2020. Understanding the relationship between direct experience and risk perception of natural hazards. Risk Analysis. 40(10):2057-2070. https://doi.org/10.1111/risa.13526
    Google ScholarOpenURLPubMed
  13. ↵
    Bubeck P, Botzen WJW, Aerts JCJH. 2012.A review of risk perceptions and other factors that influence flood mitigation behavior. Risk Analysis. 32(9):1481-1495. https://doi.org/10.1111/j.1539-6924.2011.01783.x
    Google ScholarOpenURLCrossRefPubMed
  14. ↵
    California Department of Insurance. 2024 December 30. Commissioner Lara issues landmark regulation to expand insurance access for Californians amid growing climate risks. Sacramento (CA, USA): California Department of Insurance. https://www.insurance.ca.gov/0400-news/0100-press-releases/2024/release065-2024.cfm
    Google Scholar
  15. ↵
    Calvani S, Paoloni R, Foderi C, Frassinelli N, Kirschner JA, Menini A, Galeotti G, Neri F, Marchi E. 2025. Wildfire risk perception survey: Insights from local communities in Tuscany, Italy. Fire Ecology. 21:38. https://doi.org/10.1186/s42408-025-00380-5
    Google ScholarOpenURLPubMed
  16. ↵
    Carroll MS, Kumagai Y, Daniels SE, Bliss JC, Edwards JA. 2004. Causal reasoning processes of people affected by wildfire: Implications for agency-community interactions and communication strategies. Western Journal of Applied Forestry. 19(3):184-194. https://doi.org/10.1093/wjaf/19.3.184
    Google ScholarOpenURL
  17. ↵
    Champ PA, Brenkert-Smith H. 2016. Is seeing believing? Perceptions of wildfire risk over time. Risk Analysis. 36(4):816-830. https://doi.org/10.1111/risa.12465
    Google ScholarOpenURLPubMed
  18. ↵
    Covello VT. 2021. An overview of risk communication. In: Covello VT. Communicating in risk, crisis, and high stress situations: Evidence-based strategies and practice. Hoboken (NJ, USA): Wiley. p. 33-67. https://doi.org/10.1002/9781119081753.ch3
    Google Scholar
  19. ↵
    Cox J. 2024 May 18. Trees at risk amid state insurance crisis. Bakersfield (CA, USA): The Bakersfield Californian. https://www.bakersfield.com/news/trees-at-risk-amid-state-insurance-crisis/article_e74520de-1546-11ef-add9-4b3e14d95448.html
    Google Scholar
  20. ↵
    Crowley K. 2023 July 19. Another company avoids risky Florida home insurance policies: Here’s what caused the crisis. Tysons (VA, USA): USA Today. https://www.usatoday.com/story/money/personalfinance/2023/07/19/florida-home-insurance-aaa-farmers-policy-reduction/70427062007
    Google Scholar
  21. ↵
    Darmiento L. 2024 April 19. California exodus of home insurance companies continues. Los Angeles (CA, USA): LA Times. https://www.latimes.comZbusiness/story/2024-04-19/california-exodus-of-home-insurance-companies-continues
    Google Scholar
  22. ↵
    Darrow M. 2024 August 13. San Carlos family says insurer dropped them without warning after aerial photos. Washington (DC, USA): CBS News. https://www.cbsnews.com/sanfrancisco/news/san-carlos-family-says-home-insurer-dropped-them-with-no-warning-over-oak-tree
    Google Scholar
  23. ↵
    Dash N, Gladwin H. 2007. Evacuation decision making and behavioral responses: Individual and household. Natural Hazards Review. 8(3):69-77. https://doi.org/10.1061/(ASCE)1527-6988(2007)8:3(69)
    Google ScholarOpenURLCrossRef
  24. ↵
    DeVellis RF, Thorpe CT. 2021. Scale development: Theory and applications. 5th Ed. Thousand Oaks (CA, USA): SAGE Publications Inc. 320 p.
    Google Scholar
  25. ↵
    Dewitz J. 2023. National Land Cover Database (NLCD) 2021 products. Reston (VA, USA): US Geological Survey. [USGS data release]. https://doi.org/10.5066/P9JZ7AO3
    Google Scholar
  26. ↵
    Donnellan A. 2025 February 26. Insurers will struggle to dodge climate change tab. London (United Kingdom): Reuters. https://www.reuters.com/breakingviews/insurers-will-struggle-dodge-climate-change-tab-2025-02-26
    Google Scholar
  27. ↵
    Eaglesham J. 2024 April 6. Insurers are spying on your home from the sky. New York (NY, USA): The Wall Street Journal. https://www.wsj.com/real-estate/home-insurance-aerial-images-37a18b16
    Google Scholar
  28. ↵
    Earle TC. 2010. Trust in risk management: A model-based review of empirical research. Risk Analysis. 30(4):541-574. https://doi.org/10.1111/j.1539-6924.2010.01398.x
    Google ScholarOpenURLCrossRefPubMedWeb of Science
  29. ↵
    ESRI. 2025a. ArcGIS Pro Desktop. [computer software]. Redlands (CA, USA): Esri. https://www.esri.com/en-us/arcgis/products/arcgis-desktop/overview
    Google Scholar
  30. ↵
    ESRI. 2025b. ArcGIS Living Atlas of the World. [computer software]. Redlands (CA, USA): Esri. https://livingatlas.arcgis.com/en/home
    Google Scholar
  31. ↵
    Fischer AP., Kline JD., Ager AA., Charnley S., Olsen KA. 2014. Objective and perceived wildfire risk and its influence on private forest landowners’ fuel reduction activities in Oregon’s (USA) ponderosa pine ecoregion. International Journal of Wildland Fire. 23(1):143-153. https://doi.org/10.1071/WF12164
    Google ScholarOpenURL
  32. ↵
    Flynn J., Slovic P., Mertz CK. 1994. Gender, race, and perception of environmental health risks. Risk Analysis. 14(6):1101-1108. https://doi.org/10.1111/j.1539-6924.1994.tb00082.x
    Google ScholarOpenURLCrossRefPubMedWeb of Science
  33. ↵
    Fox-Glassman KT., Weber EU. 2016. What makes risk acceptable? Revisiting the 1978 psychological dimensions of risk. Journal ofMathematical Psychology. 75:157–169. https://doi.org/10.1016/j.jmp.2016.05.003
    Google ScholarOpenURL
  34. ↵
    Frazier R. 2021 October 19. California’s ban on climate-informed models for wildfire insurance premiums. Berkeley (CA, USA): Ecology Law Quarterly. https://www.ecologylawquarterly.org/currents/californias-ban-on-climate-informed-models-for-wildfire-insurance-premiums
    Google Scholar
  35. ↵
    Gan J., Jarrett A., Gaither CJ. 2014. Wildfire risk adaptation: Propensity of forestland owners to purchase wildfire insurance in the Southern United States. Canadian Journal of Forest Research. 44(11):1376-1382. https://doi.org/10.1139/cjfr-2014-0301
    Google ScholarOpenURL
  36. ↵
    Gauldin M., Gordon J., Brodbeck A. 2025a. Fortified landscapes: A conceptual approach for how homeowners can prepare for extreme weather. Athens (GA, USA): University of Georgia Warnell School of Forestry and Natural Resources. WSFNR -25-22A. https://resources.ipmcenters.org/view/resource.cfm?rid=60709
    Google Scholar
  37. ↵
    Gauldin M., Gordon J., Brodbeck A., Calabria J. 2025b. Exploring trust and communication between insurers, arborists, and homeowners. Urban Forestry & Urban Greening. 113:129003. https://doi.org/10.1016/j.ufug.2025.129003
    Google ScholarOpenURL
  38. ↵
    Gordon JS., Matarrita-Cascante D., Stedman RC., Luloff AE. 2010. Wildfire perception and community change. Rural Sociology. 75(3):455-477. https://doi.org/10.1111/j.1549-0831.2010.00021.x
    Google ScholarOpenURL
  39. ↵
    Gray I. 2021. Hazardous simulations: Pricing climate risk in US coastal insurance markets. Economy and Society. 50(2):196-223. https://doi.org/10.1080/03085147.2020.1853358
    Google ScholarOpenURL
  40. ↵
    Gupta A, Venkataraman S. 2024. Insurance and climate change. Current Opinion in Environmental Sustainability. 67:101412. https://doi.org/10.1016/j.cosust.2023.101412
    Google ScholarOpenURL
  41. ↵
    Gustafsod PE. 1998. Gender differences in risk perception: Theoretical and methodological perspectives. Risk Analysis. 18(6): 805-811. https://doi.org/10.1111/j.1539-6924.1998.tb01123.x
    Google Scholar
  42. ↵
    Hall A, McLennan J, Marques MD, Bearman C. 2022. Conceptualising and measuring householder bushfire (wildfire) risk perception: The householder bushfire risk perception scale (HBRPS-4). International Journal of Disaster Risk Reduction. 67:102667. https://doi.org/10.1016/j.ijdrr.2021.102667
    Google ScholarOpenURL
  43. ↵
    Hoeppe P. 2016. Trends in weather related disasters—Consequences for insurers and society. Weather and Climate Extremes. 11:70-79. https://doi.org/10.1016/j.wace.2015.10.002
    Google ScholarOpenURLCrossRef
  44. ↵
    Hofmann A, Sattarhoff C. 2023. Underwriting cycles in propertycasualty insurance: The impact of catastrophic events. Risks. 11(4):75. https://doi.org/10.3390/risks11040075
    Google ScholarOpenURL
  45. ↵
    Honeycutt L. 2024 June 28. Experts say despite tightening underwriting, plenty of funding options are still available. Clearwater (FL, USA): Tampa Bay Business and Wealth. https://tbbwmag.com/2024/06/28/experts-say-despite-tightening-underwriting-plenty-of-funding-options-are-still-available
    Google Scholar
  46. ↵
    Insurance Institute for Business & Home Safety (IBHS). 2024. Fortified: A program of IBHS. Richburg (SC, USA): Insurance Institute for Business & Home Safety. https://fortifiedhome.org
    Google Scholar
  47. ↵
    Jansen SJT, Hoekstra JSCM, Boumeester HJFM. 2017. The impact of earthquakes on the intention to move: Fight or flight? Journal of Envionmental Psychology. 54:38-;49. https://doi.org/10.1016/j.jenvp.2017.09.006
    Google ScholarOpenURL
  48. ↵
    Kasperson RE, Webler T, Ram B, Sutton J. 2022. The social amplification of risk framework: New perspectives. Risk Analysis. 42(7):1367-1380. https://doi.org/10.1111/risa.13926
    Google ScholarOpenURLPubMed
  49. ↵
    Keith MG, Tay L, Harms PD. 2017. Systems perspective of Amazon Mechanical Turk for organizational research: Review and recommendations. Frontiers in Psychology. 8:1359. https://doi.org/10.3389/fpsyg.2017.01359
    Google ScholarOpenURLPubMed
  50. ↵
    Klein RW 2009. Hurricane risk and the regulation of property insurance markets. Atlanta (GA, USA): Center for RMI Research, Georgia State University. 88 p.
    Google Scholar
  51. ↵
    Klein RW, Koeser AK, Hauer RJ, Hansen G, Escobedo FJ. 2019. Risk assessment and risk perception of trees: A review of literature relating to arboriculture and urban forestry. Arboriculture & Urban Forestry. 45(1):26-38. https://doi.org/10.48044/jauf.2019.003
    Google ScholarOpenURL
  52. ↵
    Koeser AK, Smiley ET. 2017. Impact of assessor on tree risk assessment ratings and prescribed mitigation measures. Urban Forestry & Urban Greening. 24:109-115. https://doi.org/10.1016/j.ufug.2017.03.027
    Google ScholarOpenURL
  53. ↵
    Kousky C. 2017. Disasters as learning experiences or disasters as policy opportunities? Examining flood insurance purchases after hurricanes. Risk Analysis. 37(3):517-530. https://doi.org/10.1111/risa.12646
    Google ScholarOpenURLPubMed
  54. ↵
    Kowalski RM, Deas N, Britt N, Richardson E, Finnell S, Evans K., Carroll H, Cook A, Radovic E, Huyck T, Parise I, Robbins C, Chitty H, Catanzaro S 2023. Protection motivation theory and intentions to receive the COVID-19 vaccine. Health Promotion Practice. 24(3):465-470. https://doi.org/10.1177/15248399211070807
    Google ScholarOpenURLPubMed
  55. ↵
    Lee TM., Markowitz EM., Howe PD., Ko CY., Leiserowitz AA. 2015. Predictors of public climate change awareness and risk perception around the world. Nature Climate Change. 5:1014-1020. https://doi.org/10.1038/nclimate2728
    Google ScholarOpenURL
  56. ↵
    Lidskog R., Johansson J., Sjödin D. 2019. Wildfire, responsibility and trust: Public understanding of Sweden’s largest wildfire. Scandinavian Journal of Forest Research. 33(4):319-328. https://doi.org/10.1080/02827581.2019.1598483
    Google ScholarOpenURL
  57. ↵
    Lindell MK., Hwang SN. 2008. Households’ perceived personal risk and responses in a multihazard environment. Risk Analysis. 28(2):539-556. https://doi.org/10.1111/j.1539-6924.2008.01032.x
    Google ScholarOpenURLCrossRefPubMedWeb of Science
  58. ↵
    Lujala P., Lein H., Rød JK. 2015. Climate change, natural hazards, and risk perception: The role of proximity and personal experience. Local Environment. 20(4):489-509. https://doi.org/10.1080/13549839.2014.887666
    Google ScholarOpenURL
  59. ↵
    MacCallum RC., Zhang S., Preacher KJ., Rucker DD. 2002. On the practice of dichotomization of quantitative variables. Psychological Methods. 7(1):19-40. https://doi.org/10.1037/1082-989x.7.1.19
    Google ScholarOpenURLCrossRefPubMedWeb of Science
  60. ↵
    Martin WE., Martin IM., Brian K. 2009. The role of risk perceptions in the risk mitigation process: The case of wildfire in high risk communities. Journal of Environmental Management. 91(2):489-498.
    Google ScholarOpenURLCrossRefPubMedWeb of Science
  61. ↵
    Mayer A., Shelley TO., Chiricos T., Gertz M. 2017. Environmental risk exposure, risk perception, political ideology and support for climate policy. Sociological Focus. 50(4):309-328. https://doi.org/10.1080/00380237.2017.1312855
    Google ScholarOpenURL
  62. ↵
    McCaffrey S. 2004. Thinking of wildfire as a natural hazard. Society & Natural Resources. 17(6):509-516. https://doi.org/10.1080/08941920490452445
    Google ScholarOpenURL
  63. ↵
    McGee TK., McFarlane BL., Varghese J. 2009. An examination of the influence of hazard experience on wildfire risk perceptions and adoption of mitigation measures. Society and Natural Resources. 22(4):308-323. https://doi.org/10.1080/08941920801910765
    Google ScholarOpenURL
  64. ↵
    Meldrum JR., Champ PA., Brenkert-Smith H., Warziniack T., Barth CM., Falk LC. 2015. Understanding gaps between the risk perceptions of wildland-urban interface (WUI) residents and wildfire professionals. Risk Analysis. 35(9):1746-1761. https://doi.org/10.1111/risa.12370
    Google ScholarOpenURLPubMed
  65. ↵
    National Fire Protection Association (NFPA). 2024. Firewise USA. Quincy (MA, USA): National Fire Protection Association. https://www.nfpa.org/education-and-research/wildfire/firewise-usa
    Google Scholar
  66. ↵
    National Hurricane Center and Central Pacific Hurricane Center. 2024. Tropical cyclone climatology. Washington (DC, USA): National Oceanic and Atmospheric Administration. https://www.nhc.noaa.gov/climo/#:~:text=The%20official%20hurricane%20season%20for,%2DAugust%20and%20mid%2DOctober
    Google Scholar
  67. ↵
    National Oceanic and Atmospheric Administration (NOAA). 2024. Continental United States hurricane impacts/landfalls 1851-2023. Washington (DC, USA): National Oceanic and Atmospheric Administration. https://www.aoml.noaa.gov/hrd/hurdat/All_U.S._Hurricanes.html
    Google Scholar
  68. ↵
    Paolacci G, Chandler J, Ipeirotis PG. 2010. Running experiments on Amazon Mechanical Turk. Judgment and Decision Making. 5(5):411-419. https://doi.org/10.1017/S1930297500002205
    Google ScholarOpenURL
  69. ↵
    Peacock WG. 2003. Hurricane mitigation status and factors influencing mitigation status among Florida’s single-family homeowners. Natural Hazards Review. 4(3):149-158. https://doi.org/10.1061/(ASCE)1527-6988(2003)4:3(149)
    Google ScholarOpenURLCrossRefWeb of Science
  70. ↵
    Peters E, Västfjäll D, Gärling T, Slovic P. 2006. Affect and decision making: A “hot” topic. Journal of Behavioral Decision Making. 19(2):79-85. https://doi.org/10.1002/bdm.528
    Google ScholarOpenURL
  71. ↵
    Picou JS, Marshall BK, Gill DA. 2004. Disaster, litigation, and the corrosive community. Social Forces. 82(4):1493-1522. https://doi.org/10.1353/sof.2004.0091
    Google ScholarOpenURLCrossRefWeb of Science
  72. ↵
    Posit. 2024. RStudio: Integrated Development Environment (IDE) for R. Boston (MA, USA): Posit. [computer software]. https://posit.co/products/open-source/rstudio
    Google Scholar
  73. ↵
    Qualtrics. 2025. Qualtrics. Seattle (WA, USA): Qualtrics. [computer software]. https://www.qualtrics.com
    Google Scholar
  74. ↵
    Rawlinson K. 2025 October 30. Insurers calling for trees to be felled as cheap fix for subsidence, say critics. Kings Place, London (United Kingdom): The Guardian. https://www.theguardian.com/environment/2025/oct/30/insurers-calling-for-trees-to-be-felled-as-cheap-fix-for-subsidence-say-critics
    Google Scholar
  75. ↵
    Recamara J. 2025 Juy 7. Direct Line withdraws plans to fell six trees. London (United Kingdom): Insurance Business. https://www.insurancebusinessmag.com/uk/news/breaking-news/direct-line-withdraws-plans-to-fell-six-trees-541584.aspx
    Google Scholar
  76. ↵
    Rhodes N, Pivik K. 2011. Age and gender differences in risky driving: The roles of positive affect and risk perception. Accident Analysis & Prevention. 43(3):923-931. https://doi.org/10.1016/j.aap.2010.11.015
    Google ScholarOpenURLCrossRefPubMed
  77. ↵
    Robinson PJ, Botzen WJW. 2019. Economic experiments, hypothetical surveys and market data studies of insurance demand against low-probability/high-impact risks: A systematic review of designs, theoretical insights and determinants of demand. Journal of Economic Surveys. 33(5):1493-1530. https://doi.org/10.1111/joes.12332
    Google ScholarOpenURL
  78. ↵
    Rochefort C, Baldwin AS, Tiro J, Bowen ME. 2020. Evaluating the validity of the Risk Perception Survey for Developing Diabetes Scale in a Safety-Net clinic population of English and Spanish speakers. The Diabetes Educator. 46(1):73-82. https://doi.org/10.1177/0145721719889068
    Google ScholarOpenURLPubMed
  79. ↵
    Rosi A, van Vugt FT, Lecce S, Ceccato I, Vallarino M., Rapisarda F, Vecchi T, Cavallini E. 2021. Risk perception in a real-world situation (COVID-19): How it changes from 18 to 87 years old. Frontiers in Psychology. 12:646558. https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2021.646558
    Google ScholarOpenURLPubMed
  80. ↵
    Ryan RL. 2010. Local residents’ preferences and attitudes toward creating defensible space against wildfire in the Northeast pine barrens. Landscape Journal. 29(2):199-214. https://doi.org/10.3368/lj.29.2.199
    Google ScholarOpenURLCrossRefAbstract/Full Text
  81. ↵
    Siegrist M. 2021. Trust and risk perception: A critical review of the literature. Risk Analysis. 41(3):480-490. https://doi.org/10.1111/risa.13325
    Google ScholarOpenURLCrossRefPubMed
  82. ↵
    Skagerlund K, Forsblad M, Slovic P, Västfjäll D. 2020. The affect heuristic and risk perception—Stability across elicitation methods and individual cognitive abilities. Frontiers in Psychology. 11:970. https://doi.org/10.3389/fpsyg.2020.00970
    Google ScholarOpenURLPubMed
  83. ↵
    Slovic P. 1992. Perception of risk: Reflections on the psychometric paradigm. In: Krimsky S, Goldin D, editors. Social theories of risk. New York (NY, USA): Praeger. p. 117-152.
    Google Scholar
  84. ↵
    Spence A, Poortinga W, Pidgeon N. 2012. The psychological distance of climate change. Risk Analysis. 32(6):957-972. https://doi.org/10.1111/j.1539-6924.2011.01695.x
    Google ScholarOpenURLCrossRefPubMedWeb of Science
  85. ↵
    Swiss Re Institute. 2021 December 14. Global insured catastrophe losses rise to USD 112 billion in 2021, the fourth highest on record, Swiss Re Institute estimates. Zurich (Switzerland): Swiss Re Press Releases. https://www.swissre.com/media/press-release/nr-20211214-sigma-full-year-2021-preliminary-natcat-loss-estimates.html
    Google Scholar
  86. ↵
    Taber KS. 2018. The use of Cronbach’s Alpha when developing and reporting research instruments in science education. Research in Science Education. 48:1273-1296. https://doi.org/10.1007/s11165-016-9602-2
    Google ScholarOpenURLCrossRef
  87. ↵
    Tavakol M, Dennick R. 2011. Making sense of Cronbach’s alpha. International Journal of Medical Education. 2:53-55. https://doi.org/10.5116/ijme.4dfb.8dfd
    Google ScholarOpenURLCrossRefPubMed
  88. ↵
    Thomas KA, Clifford S. 2017. Validity and Mechanical Turk: An assessment of exclusion methods and interactive experiments. Computers in Human Behavior. 77:184-197. https://doi.org/10.1016/j.chb.2017.08.038
    Google ScholarOpenURL
  89. ↵
    Thomasson E. 2025 March 26. Australia and NZ face home insurance crisis due to climate, experts warn. Old Lyme (CT, USA): Green Central Banking. https://greencentralbanking.com/2025/03/26/australia-and-nz-face-home-insurance-crisis-due-to-climate-experts-warn
    Google Scholar
  90. ↵
    Torpy B. 2024 October 30. TORPY: Drone data is costing homeowners their insurance. Here’s how. Atlanta (GA, USA): The Atlanta Journal-Constitution. https://www.ajc.com/opinion/columnists/torpy-insurers-spy-cams-in-the-sky-cause-lots-of-tree-chopping/LKYZWRVY7ZCB5IKKJ5GOOIW44Y
    Google Scholar
  91. ↵
    Tree Care Industry Association (TCIA). 2023. ANSI A300 tree care standards. Manchester (NH, USA): Tree Care Industry Association, Inc. https://treecareindustryassociation.org/business-support/ansi-a300-standards
    Google Scholar
  92. ↵
    van der Linden S. 2015. The social-psychological determinants of climate change risk perceptions: Towards a comprehensive model. Journal of Environmental Psychology. 41:112-124. https://doi.org/10.1016/j.jenvp.2014.11.012
    Google ScholarOpenURLCrossRef
  93. ↵
    van der Linden S. 2017. Determinants and measurement of climate change risk perception, worry, and concern. In: Nisbet MC, Schafer M, Markowitz E, Ho S, O’Neill S, Thaker J, editors. The Oxford encyclopedia of climate change communication. Oxford (United Kingdom): Oxford University Press. 53 p. https://doi.org/10.2139/ssrn.2953631
    Google Scholar
  94. ↵
    Viscusi WK. 1991. Age variations in risk perceptions and smoking decisions. The Review of Economics and Statistics. 73(4): 577-588. https://doi.org/10.2307/2109396
    Google ScholarOpenURLCrossRefWeb of Science
  95. ↵
    Wachinger G, Renn O, Begg C, Kuhlicke C. 2013. The risk perception paradox—Implications for governance and communication of natural hazards. Risk Analysis. 33(6):1049-1065. https://doi.org/10.1111/j.1539-6924.2012.01942.x
    Google ScholarOpenURLCrossRefPubMed
  96. ↵
    Wagner KRH. 2022. Designing insurance for climate change. Nature Climate Change. 12:1070-1072. https://doi.org/10.1038/s41558-022-01514-2
    Google ScholarOpenURL
  97. ↵
    Xue W, Hine DW, Loi NM, Thorsteinsson EB, Phillips WJ. 2014. Cultural worldviews and environmental risk perceptions: A meta-analysis. Journal of Environmental Psychology. 40:249-258. https://doi.org/10.1016/j.jenvp.2014.07.002
    Google ScholarOpenURLCrossRef
  98. ↵
    Zia A, Rana IA, Arshad HSH, Khalid Z, Nawaz A 2023. Monsoon flood risks in urban areas of Pakistan: A way forward for risk reduction and adaptation planning. Journal of Environmental Management. 336:117652. https://doi.org/10.1016/j.jenvman.2023.117652
    Google ScholarOpenURLPubMed
Loading
Loading
Loading
PDF
Back to top

In this Issue

Arboriculture & Urban Forestry (AUF)

Arboriculture & Urban Forestry (AUF)

Vol. 52, Issue 4

July 2026

Table of Contents Table of Contents (PDF) Index by author

Jump to section

Keywords

Crowdsourcing, HBRPS-4, Risk Perception, Risk Scale, Underwriting Trees