Effective disaster risk governance depends heavily on public trust. Trust in authorities encourages compliance with evacuation orders, acceptance of risk reduction policies, and cooperation in community preparedness programs. Prior research in public policy and marketing underscores that transparency and honesty from institutions can build this trust. For instance, Grimmelikhuijsen et al. (2013) and Meijer et al. (2012) observed that increasing transparency tends to bolster public trust in authorities’ actions. In contrast, lack of transparency or perceived deception can quickly breed suspicion. When officials fail to clearly account for disaster funds or when corruption scandals emerge, citizens may react with anger, withdrawal, or protest. Grimmelikhuijsen et al. (2024) provide empirical evidence that targeted transparency measures (like publishing detailed spending reports) yield small but meaningful gains in citizen trust. Even if the effects are modest or contingent on context, transparency initiatives act as a “guardian of trust” by signaling accountability. Of course, such efforts must be accompanied by real performance improvements to have lasting impact. Public trust is closely tied to perceptions of government competence and fairness. People are more trusting when they see that authorities are well-prepared and that resources (aid, funds, information) are used fairly. Studies of disaster responses have found higher trust in communities that witnessed equitable aid distribution and efficient management ( Aldrich & Meyer, 2015 ; Reininger et al., 2013 ). By contrast, delays, unequal or politicized assistance, and reports of misused funds can trigger public outrage and cynicism. In this sense, trust is a form of social capital that can determine the success of emergency measures—facilitating cooperation and compliance when present, or breeding resistance and misinformation when absent. Therefore, cultivating trust is not just a nicety but a core objective of disaster governance, directly affecting outcomes. Beyond individual cases, broader patterns show that baseline trust in institutions varies by society, yet even in relatively high-trust societies it can be rapidly eroded by failures. In Türkiye, for example, the public’s trust in government institutions took decades to build but was deeply undermined after major earthquakes revealed corruption in construction and disaster funds management. After the 1999 İzmit earthquake, and again after the catastrophic February 2023 earthquakes, public confidence plummeted—particularly in vulnerable cities like Gaziantep, where seismic hazard assessments reveal significant site-specific risks ( Durukal, Erdik, & Gülkan, 2021 )—as revelations emerged about building code violations and the apparent misallocation of “earthquake taxes” intended for preparedness. Questions like “Where did the money go?” became common, reflecting a breach of trust. Once lost, such trust is difficult to restore. These events underscore that transparency and demonstrable competence are indispensable for maintaining public trust in disaster preparedness over the long term.

Budget transparency is a key pillar of governance accountability, especially in disaster preparedness funding. Citizens want to know how much is allocated to preparedness versus response, which programs are funded, and whether those funds reach their intended use. Transparent budgeting can demonstrate that leaders take preparedness seriously and have nothing to hide, thereby reinforcing trust. Lack of transparency, on the other hand, often fuels suspicions that disaster funds might be diverted or squandered. According to Grimmelikhuijsen et al. (2024) , transparency efforts can function as a “guardian of trust”, albeit with a contingent and modest effect. In their research on regulatory agencies, providing targeted information (e.g., publishing spending reports or audit results) generally increased citizen trust, though trust gains were not uniform across all groups. This implies that transparency is necessary but may not be sufficient on its own—it should be accompanied by real accountability and responsiveness. In the realm of disaster preparedness, budget accountability means not only disclosing expenditures but also demonstrating outcomes. For instance, if a government spends a certain amount on retrofitting schools against earthquakes, publicly sharing that information along with the expected risk reduction can signal competence and integrity. Visible investments in risk reduction (like strengthening infrastructure or funding community drills) reassure the public that preparedness is more than just rhetoric. However, political science research warns that leaders sometimes face skewed incentives: The electorate may reward highly visible post-disaster spending (relief aid, rebuilding projects) more than pre-disaster investments that quietly prevent harm. This myopic voting dynamic, identified in earlier studies by Healy and Malhotra (2009) and confirmed in recent work by Morvan and Paty (2024) , can discourage officials from prioritizing funding for preparedness. Voters tend to applaud relief spending that they can see in the aftermath of a disaster, but often fail to reward (or even notice) preventive spending that averts disaster impacts. As a result, politicians might be tempted to quietly reallocate preparedness funds elsewhere, confident that the public won’t realize the loss until it’s too late Strengthening budget transparency in disaster management involves several strategies. One is performance auditing—regularly evaluating and publicly reporting on how preparedness funds are used and what results they achieve. Another is participatory budgeting, where community stakeholders have a voice in how disaster preparedness monies are spent. By engaging the public in budgeting decisions, officials can both educate citizens about the trade-offs involved and gain legitimacy for those choices. If people see, for example, that a new emergency shelter was built instead of repaving a highway because of disaster risk analysis, they might be more understanding and trusting of those decisions. In summary, transparency and accountability in budgeting can illuminate the often unseen work of disaster preparedness, thereby bolstering public trust that scarce resources are being managed wisely and ethically.

Political ideology serves as a lens through which individuals interpret risk information and evaluate government actions. Conservatives, moderates, and liberals may all view the same disaster preparedness policy differently due in part to their ideological values and their baseline trust in government. Research has shown that ideology can influence how people process messages and to whom they assign blame or responsibility in crises. For example, a marketing study by Chan and Ilicic (2019) demonstrated that political ideology correlates with the strength of emotional attachment individuals form with organizations and causes. In that work, consumers with more conservative leanings showed a tendency to form stronger brand attachments (seeking security in familiar brands during uncertainty), whereas more liberal individuals were less brand-attached. By analogy, in a disaster context this suggests a person’s ideological bent might affect whether they trust government-led preparedness efforts or prefer community and private initiatives. Those who are ideologically conservative might be inclined to trust traditional authorities and institutional assurances, while those who are more liberal might exhibit skepticism and demand more grassroots or oversight mechanisms. Ideology also intersects with risk perception. Prior studies indicate that individuals with more conservative leanings sometimes downplay certain risks or place greater trust in decentralized, individual responses, whereas more liberal individuals might emphasize collective action and government responsibility in risk mitigation ( Barrios & Hochberg, 2021 ). A recent study on earthquake preparedness in Chile ( Repetto, Cordón, & Bronfman, 2022 ) found that political ideology was a significant factor in both trust and preparedness behaviors. In that survey, respondents on the right-end of the ideological spectrum (who tended to express higher trust in authorities) reported slightly higher levels of personal preparedness for earthquakes and tsunamis compared to others. This somewhat counterintuitive finding—that those more trusting of government were also engaging in preparedness actions—underscores the complex ways ideology can shape both attitudes and behaviors. It suggests that conservative individuals, despite trusting authorities, may still take action to protect themselves, possibly due to a sense of personal responsibility or higher risk awareness in that context. Meanwhile, liberal individuals, often more skeptical of authorities, might engage in preparedness from a distrust motive or a collective ethic. The key point is that worldview and prior beliefs shape how people respond to disaster risk and government directives. Understanding these nuances is crucial for campaign strategists and policymakers. Messages about disaster preparedness can be tailored to resonate with different ideological groups—for instance, a campaign in a conservative-leaning region might frame preparedness in terms of family security and self-reliance, whereas in a more liberal area it might stress social justice and protecting vulnerable communities. By aligning communication with ideological values, leaders can better build trust and motivate action across a diverse public.

We approached these questions through a mixed-method design. First, we reviewed interdisciplinary literature spanning public administration, marketing, and disaster studies to build hypotheses about trust dynamics. Next, we conducted a survey targeting residents in Türkiye, a country acutely familiar with seismic disaster risk. The survey was administered online in mid-2024, approximately one year after a series of devastating earthquakes struck the region. We obtained a sample of 486 respondents across various regions of Türkiye. Participants were recruited via a research panel to approximate national diversity in age, gender, and region, though the sample is not fully representative. Ethical research approval and informed consent were obtained prior to data collection.

The survey collected demographic data and included the following key measures: (a) Trust in Government Budget Allocation for Disaster Preparedness—a single-item rating (“To what extent do you trust that the government is using disaster preparedness funds properly?”) on a 1 (Not at all) to 5 (Completely) scale; (b) Perceived Government Transparency—a 1 - 5 scale rating agreement with “The government is transparent in how it handles the disaster preparedness budget”, where 1 = strongly disagree (not transparent) and 5 = strongly agree (highly transparent); (c) Voting Priority for Disaster Preparedness—measured by asking respondents how important disaster preparedness is to them when voting in national or local elections (1 = Not at all important, 5 = Extremely important); (d) Political Skepticism—a composite measure (α = 0.81) that captures generalized cynicism toward political figures and institutions. It includes multiple Likert-scale items assessing beliefs such as “Politicians usually lie about their true intentions”, “Most politicians are only in it for themselves”, and “Elected officials rarely care about people like me”. Together, these items reflect a broader distrust in the motives and honesty of political actors; (e) Desire for Specific Disaster Plans in Campaigns—a yes/no item asking if respondents want political candidates to present specific disaster preparedness plans (we also recorded qualitative examples, though not analyzed here in depth); (f) Belief that Authorities Learn from Past Disasters—a 1 - 5 agreement rating with “We learn from past disasters to improve future outcomes”; (g) Personal Disaster Experience—a binary indicator if the respondent has personally experienced a serious disaster (earthquake, flood, etc.) that affected their life or household; (h) Ideological Self-Identification—respondents chose a category that best described their political stance (“Conservative”,, “Moderate”, “Liberal”, “Apolitical/Not interested”, or “Other”); (i) Regional Risk Level—we recorded the respondent’s city and coded regions by objective risk (cities in historically high-risk zones vs. lower-risk areas) and by a simple urban/rural classification. Finally, we included a few behavioral questions (e.g., “Do you feel you are personally prepared for a future disaster? yes/no”). The survey instrument was developed in both English and Turkish, then translated/back-translated for accuracy in the Turkish version.

We used a combination of quantitative analyses. To compare groups (e.g., survivors vs. non-survivors; different ideologies), we applied independent-samples t-tests and one-way ANOVAs as appropriate. Where significant effects were found, we conducted Tukey’s HSD post-hoc tests to identify which specific groups differed from each other, and we calculated effect sizes (η² or Cohen’s d) to gauge practical significance. For correlational analysis, we computed Pearson correlation coefficients among key continuous variables (trust in budget, perceived transparency, skepticism, etc.) and between those measures and the voting priority variable. A correlation matrix was produced (visualized in a heatmap) to examine the direction and magnitude of relationships; we flagged correlations as significant based on p < 0.05 and noted their magnitudes. To explore predictive relationships and control for overlaps, we ran a multiple linear regression with Vote Priority as the dependent variable. Predictors entered were those we theorized could influence the prioritization of disaster preparedness in voting: Trust in Budget, Political Skepticism, Specific Messaging preference, and belief that authorities Learn From Past disasters. We reported the overall model fit (F-test and R²) and each predictor’s standardized coefficient (β) with its significance level. Lastly, to investigate regional effects, we compared respondents by region type. We used both the city-level data (identifying the top ten cities by respondent count, many of which were high-risk locales) and a broader region classification (grouping responses into “Urban”, “Disaster-prone region”, “Suburban”, and “Rural” categories). We ran ANOVAs to test for regional differences in Vote Priority and in trust measures. All statistical analyses were conducted using SPSS (v28). Throughout, a significance level of 0.05 (two-tailed) was used. Given multiple comparisons, we interpret findings with caution, emphasizing effect sizes and consistency with theory. The following section presents the results in detail, organized by our focal themes: experience effects, ideological differences, trust–engagement dynamics, and regional factors.

A citizen’s personal experience with disaster emerged as one of the most potent influences on trust in the government’s disaster preparedness spending. As expected, respondents who had directly experienced a disaster (whom we term “survivors”) exhibited far lower trust in the government’s use of the disaster budget than those with no such experience. The contrast is striking: disaster-experienced individuals overwhelmingly reported very low trust. The median trust score among survivors was 1 (the minimum on our 1 - 5 scale), with very little variation – in fact, the entire interquartile range lies at the floor of 1 in the survivor group. Fully 88% of disaster survivors answered “1 – Not at all” when asked if they trust the government’s use of the disaster preparedness budget. A few survivors gave a score of 2, appearing as minor outliers above the box in our analysis, but notably none of the survivors rated their trust higher than 2. By contrast, those without disaster experience (“non-survivors”) exhibited a broader distribution of trust ratings: the median in the non-survivor group was around 3.0 (indicating a neutral to slightly distrustful stance), and responses spanned the full range up to 5. A small minority of non-experienced respondents did express complete trust (5 out of 5), showing that without direct experience some individuals maintain faith in the system. The interquartile range for non-survivors extended roughly from 2 to 4, highlighting both much greater variability and generally higher trust than the survivor group. Statistical tests confirm that this trust gap is highly significant. An independent-samples t-test (Welch’s two-sample t-test, not assuming equal variances) comparing mean trust scores between the two groups found t = –17.50, p ≈ 2.3 × 10⁻⁴². Survivors’ mean trust was M = 1.12 (SD ≈ 0.33), while non-survivors’ mean was M = 2.82 (SD ≈ 1.31). The difference of −1.70 points on the 5-point scale corresponds to an extremely large effect size (Cohen’s d ≈ 1.99). In plain terms, those who have lived through disasters have nearly zero trust that the allocated preparedness funds are being properly used, whereas those without such experience are more ambivalent (though still leaning towards distrust on average). This lends strong support to the notion that personal disaster experience erodes public trust in the government’s disaster readiness expenditures. It is noteworthy that even the non-experienced group’s average trust (about 2.8) was below the midpoint “3” on the scale, suggesting a general skepticism in the population at large—but nothing as severe as the near-total distrust observed among survivors. Thus, while mistrust is widespread, it is especially acute among those with direct experience of disasters. This finding aligns with the idea that experiencing a traumatic event can make individuals more skeptical and vigilant about promises of safety and proper resource use. Beyond personal experience, we also explored how an individual’s perceived risk and their view of budget transparency relate to trust in preparedness efforts. Figure 1 summarizes these relationships. As the chart illustrates, respondents who believed that the budget allocation process is transparent reported higher trust in the government’s disaster preparedness spending than those who believed it is not transparent—and this held true at both low and high levels of perceived disaster risk. However, the lowest trust of all was observed among those who perceived the disaster risk to be high and felt that the budget was not transparent. In other words, high risk perception coupled with low transparency perception led to a compounded distrust. Conversely, the relatively highest trust levels were found among individuals who perceived low disaster risk and also believed the budgeting process is transparent (though notably, even this group’s mean trust did not reach the scale midpoint, reflecting overall skepticism). These patterns suggest that both risk awareness and transparency beliefs shape trust: people who feel personally safe or unconcerned about risk can maintain some trust as long as they think the government is transparent, whereas those who sense great risk will be distrustful especially if they also doubt the honesty of the budgeting. While the influence of subjective risk perception was not as dramatic as the survivor vs. non-survivor gap, it adds another layer to understanding trust dynamics. Those who perceive higher risk tend to be more critical of government preparedness efforts (perhaps because they are more attuned to what is at stake), and this effect is amplified when transparency is in doubt. Notably, despite survivors’ deep distrust, we considered whether this lack of trust translated into different behaviors or attitudes regarding disaster preparedness. One might expect, for example, that distrustful survivors either withdraw from political engagement (cynicism leading to apathy) or become more vocal about demanding preparedness (mistrust fueling activism). Our findings align more with the latter interpretation: survivors did not disengage from the issue politically—if anything, they appeared slightly more adamant that disaster preparedness should be a top priority. The mean Vote Priority (the self-reported importance of disaster preparedness when voting) was extremely high for both groups: 4.63 for survivors vs. 4.62 for non-survivors—essentially no difference. Both groups were near the ceiling of 5, indicating that almost everyone considered it a top issue, regardless of their trust level. There was also a small difference in self-reported personal preparedness actions: 51.9% of disaster survivors said they felt they were “prepared” for a future disaster (e.g., having supplies, plans, etc.), compared to 44.9% of those with no disaster experience—about a 7 percentage point gap. However, a Z-test for this proportion difference yielded z = 1.48, p = 0.14, which is not statistically significant. Thus, while there is a hint that survivors take more personal preparedness steps (perhaps due to their firsthand awareness), our data did not conclusively show a large gap in self-reported preparedness levels. What is clear is that survivors harbor deep skepticism toward government efforts, which could have implications for how they respond to official directives or initiatives in the future. Personal disaster experience emerged as a powerful factor shaping public trust. Going through a disaster is associated with an almost two-standard-deviation drop in trust in the government’s disaster-related budget use. For policymakers, this signals that the very people who have the most at stake (survivors in high-risk areas) are also the hardest to convince or reassure, likely because they have witnessed perceived failures firsthand. Rebuilding trust among this group may require substantially greater transparency and proof of improvement, as mere promises will likely fall on deaf ears. In the Discussion, we further explore strategies to address this trust deficit, such as targeted community engagement and visibly correcting past shortcomings.

Political ideology showed a subtle influence on how participants viewed government transparency in disaster preparedness. Figure 2 presents the mean transparency perception scores for respondents of different ideological self-identifications: Conservative, Moderate, Liberal, Apolitical, and Other. The transparency perception question asked respondents to rate how transparent they find the government’s actions in disaster preparedness on a 1 to 5 scale. A score of 1 indicates the view that the government is highly non-transparent/opaque, while 5 would indicate a view of high transparency. In our data, no group’s average even approached the midpoint of 3, underscoring a pervasive skepticism about transparency across the spectrum. As depicted in Figure 2 , all ideological groups reported low transparency on average, with mean values clustered roughly between about 1.3 and 1.7—well below a “neutral” 3 on the scale. In practical terms, the typical respondent, regardless of ideology, tended to answer that the government is “rarely transparent” or at best only “sometimes transparent”.

This underscores a broad perception of opacity in the government’s disaster preparedness efforts. Despite universally low transparency perceptions, there were statistically significant differences among the groups. Conservatives showed the “highest” mean transparency perception at around 1.64 (their bar in Figure 2 is the tallest, though still very low in absolute terms). Liberals, Apoliticals, and those in the “Other” category (which includes mixed or unspecified ideologies) had similar means around 1.43 - 1.45. Moderates were the most skeptical of transparency, with a mean of roughly 1.30—the lowest among the groups. A one-way ANOVA on the transparency rating by ideology confirmed that not all groups share the exact same mean perception (F(4, 481) = 4.95, p < 0.001). The effect size was modest (η² ≈ 0.04), meaning about 4% of the variance in perceived transparency can be attributed to political ideology. Post-hoc comparisons (Tukey’s HSD) revealed where the differences lay: Conservatives had significantly higher transparency perceptions than Liberals (mean difference ≈ 0.21, p ≈ 0.047) and significantly higher than Moderates (diff ≈ 0.34, p = 0.0001). These pairwise differences align with the slight uptick observed for conservatives in the figure. No other pairwise differences were statistically significant at the 0.05 level—for instance, Liberal vs. Moderate or Apolitical vs. Other comparisons showed gaps (liberals vs. moderates differed by ~0.13, liberals vs. apoliticals by ~0.02, etc.) that did not reach significance given sample sizes. Thus, the main nuance is that self-identified conservatives view the government as slightly more transparent than others do, whereas moderates, liberals, apolitical individuals, and those identifying as “other” hold even more negative views of transparency. Importantly, by any absolute standard, even conservatives’ average view was “non-transparent” (much closer to 1 than to 5 on the scale). This finding might initially seem counterintuitive, as one might expect conservatives—who often favor limited government—to distrust government transparency as much or more than liberals do. However, a possible explanation lies in the contemporary political context. In our survey setting (Türkiye), the ruling government has a conservative orientation. Thus, many conservative respondents may be supporters of the incumbent government and therefore more inclined to trust its communications or give it the benefit of the doubt. Meanwhile, moderates and liberals might be more critical or expect a higher standard of openness, leading them to rate transparency lower. Another interpretation is that conservatives’ baseline for what counts as “transparent” governance might differ—they might not require as much public disclosure to feel that a government is doing its job, whereas liberals might be more demanding of detailed disclosures and hence judge the transparency as lacking. These interpretations extend beyond our data, but they illustrate how ideology and partisan alignment can color perceptions of the same reality. It is also worth noting that conservative ideology is sometimes associated with a general trust in authority and social order, which can translate into greater trust in institutions when one’s preferred party is in power. Conversely, liberal ideology often emphasizes accountability and skepticism of authority, which might lead liberals to be especially attuned to transparency shortcomings. Our data, in effect, captured a small echo of these tendencies: conservatives weren’t highly trustful (they were still very skeptical on an absolute scale), but they were slightly less distrustful than others. Crucially, despite conservatives rating transparency a notch higher, the overall story is one of low perceived transparency across the board. Even the most “optimistic” group’s mean (about 1.64) was much closer to “highly opaque” (1) than to “fully transparent” (5). This pervasive skepticism matches reports from other countries that only a small fraction of citizens consider their government open and honest about disaster management policies. In Türkiye, for instance, broad public surveys and media discussions have indicated that many people believe information is withheld or spun for political reasons. Our findings mirror that reality, indicating a trust deficit that transcends ideological divisions. To double-check these results, we also looked at a related variable: Trust in Official Information (how much respondents trust the information and updates provided by officials regarding disasters). This measure showed a similar pattern: overall trust in official information was low (mean around 2.9/5, which is still on the distrustful side) and we found no dramatic differences by ideology, except again a mild conservative uptick. The correlation between the perceived transparency score and trust in information was positive (as expected—people who think the government is transparent also tend to trust its information) but only moderate (around r = 0.3). Interestingly, some respondents seemed to think the government was not very transparent yet still somewhat trusted the limited information provided, or vice versa. This suggests that while related, perceived transparency and trust in information are not identical constructs; some people might trust specific communications (perhaps due to the urgency or accuracy of emergency alerts) even if they generally feel kept in the dark. In conclusion, ideology had only a minor effect on views of transparency—the key takeaway is that all groups, regardless of political leaning, perceive transparency to be low in the context of disaster preparedness. The minor variation that exists (with conservatives slightly less negative) likely reflects political alignment with the current government. From a strategic standpoint, these results imply that improving transparency would be universally beneficial, as skepticism is widespread. Political leaders cannot assume their base automatically perceives transparency; even their supporters are quite skeptical in absolute terms. Therefore, addressing transparency concerns and demonstrating openness should be a cross-cutting effort, not one targeted to just one side of the aisle.

One of our research questions was how trust-related attitudes connect to political engagement with disaster issues. We approached this by examining simple correlations and then a regression model involving Vote Priority—the importance respondents assign to disaster preparedness when voting—and variables like trust in budget use, desire for specific disaster messaging from politicians, perceived learning from past disasters, and general political skepticism. Before delving into specific relationships, it is worth noting a descriptive finding: overall, people placed a very high priority on disaster preparedness as a voting issue. The mean rating on the 1 - 5 importance scale was above 4.6, and nearly 80% of the entire sample selected the maximum of “5 – Very high priority”. Virtually no one rated it below 3. This indicates a ceiling effect—an across-the-board recognition that disaster readiness is a crucial issue for voters. In such a context, detecting differences or predictors of Vote Priority is challenging because there is little variation to explain. Nonetheless, subtle patterns did emerge in relation to trust and skepticism. We produced a correlation matrix ( Figure 3 ) to see the direction and magnitude of relationships among key variables where warmer colors (toward red) indicate positive correlations and cooler colors (blue) indicate negative correlations, with the scale centered at zero.

All correlations are modest to near-zero, underscoring that no strong linear relationship emerges between these factors. For instance, greater distrust in budget allocation correlates only slightly with higher prioritization of disaster preparedness in voting (r ≈ −0.08), and higher general political skepticism shows a small negative link with vote priority (r ≈ −0.15). Similarly, believing that authorities “learn from past disasters” has a mild positive association with treating preparedness as a voting priority (r ~ +0.12). In line with the text, virtually all pairwise correlations are weak (absolute r < 0.3), reflecting a ceiling effect in vote priority and generally pervasive distrust across respondents. Focusing on the correlations involving Vote Priority, a few noteworthy (if modest) associations were observed:

We also checked Gov Transparency vs. Vote Priority (perceived transparency rating correlated with importance of issue). This yielded r ~ +0.01, essentially no relationship. This is plausible because, as we saw, almost everyone perceives low transparency; there wasn’t much variation to correlate with voting priority, and conversely, voting priority was near-maximal for almost everyone.

Most pairwise correlations among the trust-related items themselves were close to zero or modest. For instance, Trust in Budget and a related measure of perceived fund effectiveness correlated at r ~ +0.07 (a slight positive, meaning people who trust the budget also think funds are used a bit more effectively—logically consistent, but very small in magnitude). Interestingly, Political Skepticism had near-zero correlation with Trust in Budget (r ≈ 0.00). This suggests that general cynicism about politicians didn’t strongly predict one’s specific trust in the disaster budget—likely because almost everyone was distrustful of the budget, even people who weren’t uniformly skeptical otherwise. In other words, even respondents who might normally trust politicians in general had low trust in this particular domain (disaster funding), indicating a consensus distrust.

In sum, the correlation analysis suggests that no single factor strongly “drives” the prioritization of disaster preparedness in voting—largely because that priority is uniformly high. However, there are hints of relationships: generalized mistrust (skepticism) correlates with a bit less engagement on the issue, whereas specific trust deficits (low trust in the budget) correlate with slightly more engagement. This nuanced pattern implies a distinction between cynicism-induced apathy and issue-specific outrage. High general skepticism might lead some to tune out (not prioritizing any policy issues because they expect little), while distrust focused on disaster management might motivate others to demand better performance through their vote. To further parse these relationships, we ran the multiple regression described in the Methodology. The regression results showed a significant overall model (F(4, 481) = 5.47, p < 0.001), but a small R² = 0.044, meaning only ~4.4% of the variance in Vote Priority was explained by the included factors. This low R² is not surprising given the ceiling effect on Vote Priority—with so many respondents rating it as very important, there was limited variance to explain. It also indicates that other unmeasured factors (or simply a uniform concern across the board) account for the vast majority of why people rank this issue so highly. Within the regression model, two predictors emerged with significant coefficients. Political Skepticism had a negative coefficient (β ≈ −0.059, p = 0.001), aligning with the correlation: higher skepticism predicts slightly lower priority on this issue, controlling for other variables. Specifically, for every one-step increase in skepticism (on the 1 - 5 scale), the Vote Priority rating dropped by ~0.06 (on the 1 - 5 scale), holding other factors constant. In practical terms, a very skeptical person (5 on skepticism) might rate the issue a tiny bit lower than a moderately skeptical person (4 on skepticism), all else equal—a small effect, but in a consistent direction. Belief in Learning from Past had a positive coefficient (β ≈ +0.047, p = 0.010): those who think that authorities do learn and improve after disasters assign slightly more importance to preparedness in voting, as if they feel their vote can encourage continued improvement. For example, someone who “strongly agrees” that lessons are learned might rate the voting priority perhaps 0.1 - 0.2 points higher than someone who “strongly disagrees”, according to the model. Meanwhile, Trust in Budget had a negative coefficient (β ≈ −0.037) that was marginally significant (p = 0.076), mirroring the weak negative trend from the correlation: those with higher trust tended to give slightly less priority to the issue, whereas those with low trust were more inclined to prioritize it. In the regression, this was only a borderline effect (we might call it a trend rather than a definitive finding). The Specific Messaging preference was not significant at all (β ≈ −0.005, p = 0.77), indicating it had no unique predictive value when other variables were accounted for ( Figure 4 ). The diagram below illustrates the multiple linear regression findings, showing how four key independent variables relate to Vote Priority (the self-reported importance placed on disaster preparedness when voting). Each arrow corresponds to a predictor’s standardized beta coefficient (β), with significant relationships noted by asterisks (p < 0.05, p < 0.01). Political Skepticism had a negative effect (β ≈ −0.06, p = 0.001), indicating that more cynical individuals assign slightly lower priority to disaster preparedness in elections. Conversely, the belief that “we learn from past disasters to improve” showed a positive effect (β ≈ +0.05, p = 0.010)—suggesting a bit of higher priority among those hopeful about learning and improvement. Trust in Budget Use exhibited a small negative coefficient (β ≈ −0.04, p ≈ 0.08), mirroring a weak trend that those who trust the government’s disaster spending slightly less often treat preparedness as a voting priority. Lastly, the preference for Specific Messaging from politicians had essentially no unique effect (β ≈ −0.01, n.s.), once other factors were accounted for. These modest coefficients align with the text’s finding of a low overall R² (~0.04) for the model—meaning that individual differences in trust or skepticism explain only about 4% of the variance in an issue that almost everyone rated as “very important”. Thus, our analysis did not find very strong predictors of treating disaster preparedness as a voting priority—largely because almost everyone rated it as quite important. The lack of variation (a ceiling effect) means there was little room for predictors to explain differences. In statistical terms, with so many giving the highest possible importance, variance was limited and our model’s R² remained low. Still, the fact that political skepticism slightly dampened the priority could reflect a segment of the population that is so disenchanted that they’ve “checked out” on expecting solutions (hence not prioritizing it politically). Meanwhile, those who have not completely lost faith (they think learning and improvement is possible) seem to channel that into making the issue a priority, perhaps hoping to elect competent stewards. In other words, a bit of hope in the system corresponds with treating the issue as even more critical—possibly because these individuals believe that good governance can make a difference in outcomes. Descriptively, it’s enlightening to note just how high the prioritization was: nearly 80% of our sample answered 5 (“very high priority”) for the voting importance of disaster preparedness, and virtually none answered below 3. This unanimity reflects the salience of disaster issues in the collective consciousness, likely heightened by recent catastrophic events in the national narrative. In such a context, differences in degree of priority are subtle. Even so, our findings collectively imply that where differences exist, they suggest that distrust (especially specific distrust in the current government’s handling) is associated with a greater push to emphasize the issue in elections. People who think “something is wrong” are logically more likely to demand change or specific actions from candidates. Indeed, the slight negative correlation between trust in government and priority placed on the issue can be interpreted as evidence that those who trust the status quo on preparedness are content (and thus see no need to make it an election issue), whereas those who don’t trust it want it elevated on the campaign agenda to force improvements. Additionally, we examined whether regional context influenced Vote Priority—the idea of a “regional sensitivity” effect. Neither correlation nor ANOVA found region type to be a significant predictor of vote priority (the ANOVA for differences by broad region categories had p = 0.44). Essentially, people across the country—whether in high-risk zones or not—all rated disaster preparedness as a very high priority to a similar extent. This could be due to widespread media coverage making disasters a national concern, or shared empathy and awareness that “it could happen here too”. We explore this further in the next section on regional differences. The correlation and regression analyses suggest that trust-related attitudes have at best modest linear relationships with the political prioritization of disasters, largely because the priority is uniformly high. However, the data hint that low trust in government (budget distrust) does not lead to apathy; if anything, it correlates with a motivation to make it a political issue (though our ability to detect that was limited by the ceiling effect). The one factor clearly associated with reduced prioritization was general political cynicism. These nuances are important for campaign strategists – motivating an already concerned public likely requires addressing their trust issues so that cynicism doesn’t turn into disengagement.

A key question was whether those living in objectively high-risk regions (especially major cities prone to earthquakes) exhibit different attitudes compared to those in lower-risk areas. After examining the national-level findings, we wanted to see if local context—being in a disaster hot zone versus a calmer area—made a measurable difference in trust or prioritization. Do residents of, say, Istanbul or Kahramanmaraş (cities with known seismic risk or recent disaster impact) prioritize preparedness more, or trust the government less, than residents elsewhere? We first compared respondents across broad regional risk levels. For this analysis, we classified the survey sample’s cities into three categories based on known seismic risk and recent disaster history: High-risk cities, Moderate-risk cities, and Low-risk cities. Figure 4 shows the average Vote Priority (importance of disaster preparedness in voting, on the 1 - 5 scale) for each of these three risk-level categories. The differences were minimal—all categories were extremely high. High-risk cities (e.g., those in seismic zones) showed an average around 4.7; moderate-risk cities were about 4.6; low-risk cities around 4.5. Statistically, an ANOVA found no significant differences among the three groups (p > 0.40). In practical terms, people treated the issue as critical regardless of whether their region was officially high-risk or not. While the broad risk categories did not show significant differences, we conducted a more granular city-level analysis focusing on major cities in the sample. Many of these cities are known for high disaster risk or have recently experienced disasters. They include places like Kahramanmaraş (epicenter of a major 2023 earthquake), Istanbul (a megacity with significant seismic risk), Düzce (site of past earthquakes), Erzurum, Trabzon, Kütahya, Van, Mersin, İzmir, and Bursa. We observed that all these cities have extremely high average scores, generally between 4.5 and 5. Kahramanmaraş tops the list with an average nearly at 5.0 (virtually every respondent there said it’s a top priority when voting). Istanbul and Düzce also show averages above 4.8. Even the “lowest” among the top ten—for instance, Bursa in this figure—is still around 4.4 to 4.5, which is extremely high on the scale. Residents of earthquake-stricken or high-risk cities are essentially in agreement that this issue cannot be ignored by their leaders. This matches expectations: communities that have felt devastation firsthand (or know they are likely to) have zero complacency on the matter. For example, Kahramanmaraş respondents, having lived through a catastrophic quake, almost unanimously rated preparedness as a top priority. Istanbul, which awaits a potentially severe earthquake in the future, also has its residents highly attuned to the issue. Interestingly, when we broaden beyond these top cities and use our broader region classifications (as mentioned above), we find that all regions treated preparedness as a high priority. For instance, averaging by region type: Urban region mean = 4.68, Disaster-prone region mean = 4.65, Suburban region mean = 4.61, Rural region mean = 4.57. The differences between, say, an officially designated “Disaster-prone” region and a “Rural” region was only 0.11 on the 5-point scale. As noted, our ANOVA showed no significant regional effect on Vote Priority (p = 0.44). This suggests that even people in less hazard-exposed areas recognize the importance of preparedness—possibly due to national solidarity, extensive media coverage of disasters, or the realization that no area is entirely risk-free. There is a subtle hint that urban and high-risk city respondents gave numerically the highest scores (urban residents, many of whom are in cities like Istanbul or İzmir, had the highest mean of ~4.68; those in the specific “disaster-prone region” category were also ~4.65). While these differences were not statistically distinct, the practical reality is that in places that recently suffered disasters, virtually everyone marks it as a top issue (often with very little variance, as seen in Kahramanmaraş where almost all respondents chose 5). Meanwhile, in rural or less directly affected areas, there might be a few more people who rated it a 4 instead of 5, explaining a tiny drop in the average. We also examined regional differences in the trust variables. One might expect that those in high-risk or disaster-affected regions trust the government less, owing to having seen failures up close. Indeed, respondents in the “Disaster-prone” regional category had the lowest mean trust in budget (mean ~1.69 out of 5), compared to urban residents (~1.90 out of 5). This suggests a similar pattern: those in the thick of it (high-risk areas) are slightly more distrustful. However, this difference was not statistically significant in our sample (ANOVA p ≈ 0.33 for trust by region). It appears that distrust is widespread geographically, not just confined to the hardest-hit areas. Similarly, perceived transparency was very low across all regions; interestingly, urban respondents had marginally higher transparency perceptions than others (perhaps reflecting closer proximity to government and media sources), but again the differences were small. In terms of self-reported preparedness behaviors, we found no major distinctions by region: residents of rural vs. urban vs. disaster-prone regions all had roughly 47% - 52% saying they feel prepared themselves, with no clear trend that high-risk area residents felt more prepared. This indicates that being in a high-risk area doesn’t necessarily translate to feeling personally prepared—perhaps due to resource or knowledge gaps even in those areas. So, what do these results mean for “regional sensitivity”? In one sense, citizens in disaster-affected cities are highly sensitive to and unified on this issue—as our figures and data confirm. But a crucial finding is that citizens everywhere rate the issue as critically important and trust as low. The whole country, not just the disaster zones, seems to be paying attention. This could reflect a general alarm after seeing disasters strike (for instance, even those in suburban towns saw the chaos in major cities on TV and want better national preparedness). It might also reflect shared cultural memory or the fact that many people have friends/family in affected areas. From a policy perspective, this means politicians cannot afford to ignore disaster preparedness in any region; it’s not a niche local issue but a national one. If one imagined a map of the country shaded by trust levels, it would be almost uniformly low-confidence; if shaded by priority given, it would be almost uniformly high concern. Earthquake-prone cities stand out perhaps in the intensity and unanimity of their concern (acting as epicenters of public pressure), but they are far from alone. In summary, the data suggest strong national solidarity in treating disaster preparedness as a political priority. Those in high-risk cities are (as expected) among the most vocal and unanimously concerned, but these differences are more a matter of degree since even “low-risk” regions show very high concern. While our quantitative tests didn’t flag region as a significant differentiator for priorities or trust, the real-world implication is that recent disasters have made preparedness a front-and-center issue countrywide. Any regional tailoring of strategy might focus more on content (e.g., earthquake retrofitting plans in seismic zones versus flood infrastructure in flood-prone areas) rather than on convincing people of the issue’s importance—the importance is recognized everywhere. Having presented these results—the experiential trust gap, the broad skepticism with slight ideological nuances, the modest links between trust and political prioritization, and the consistent regional concern—we now turn to interpret what these patterns mean for public policy and political strategy in disaster preparedness in the conclusion.