When fear disrupts growth: Modeling the economic toll of terror on tourism

Abstract

This paper develops a dynamic Ramsey-Cass-Koopmans (RCK) model to analyse the macroeconomic impact of terrorism-induced fear on tourism and growth. We introduce a behavioural mechanism where households’ utility from tourism consumption is distorted by a time-varying security perception index, $\theta \left(t\right)$ , which declines following terrorist attacks. This fear channel reduces savings and investment, slowing capital accumulation and long-run growth. Our model also incorporates government intervention through public safety investment aimed at restoring confidence. We evaluate three scenarios: (i) one-time shock; (ii) persistent fear; and (iii) policy-driven recovery, and simulate their effects on capital, consumption, and welfare. Our results show that persistent fear leads to the greatest welfare loss, while rapid policy-driven recovery can paradoxically destabilize investment if not properly calibrated. Our paper highlights the trade-offs in post-terrorism policy design and offers new theoretical insights into how behavioural responses to insecurity can shape macroeconomic trajectories in tourism-dependent economies.

Introduction

Tourism has emerged as one of the fastest-growing and most economically significant sectors globally, accounting for over 10% of global GDP and supporting millions of jobs across both developed and developing nations. For many countries, particularly small island economies and developing regions, tourism represents a primary source of foreign exchange earnings, infrastructure investment, and socio-economic development. However, the industry is vulnerable to exogenous shocks: none more psychologically potent and economically disruptive than tourism. Acts of terror, even geographically isolated or limited in scope, can lead to disproportionate declines in international tourist arrivals due to fear-driven behavioural response. These events often result in long-lasting reputational damage, reduced investor confidence, and structural shifts in consumption patterns, particularly in sectors like hospitality, transport, and cultural services.

A large body of empirical literature has documented the adverse effects of terrorism on tourism flows. Enders and Sandler (1991) provided early evidence that tourist incidents in Spain led to significant reductions in tourist arrivals. Similar findings were documented Drakos and Kutan (2003), who documented that terrorism adversely affected regional tourism in Mediterranean countries. Neumayer (2004) used dynamic panel estimation techniques to show that political violence and terrorism have strong and persistent negative effects on international tourism across a wide set of countries. These impacts often extend beyond the immediate aftermath of attacks, as tourists adjust their perceptions of safety based on both direct threats and media portrayals.

Recent contributions have emphasized the psychological and behavioural dimensions of terrorism’s economic toll. Sönmez and Graefe (1998) found that perceived risk and safety concerns significantly influence travel behaviour, often outweighing actual statistical probabilities of harm. Frey et al. (2007) further argues that the fear induced by terrorism affects not only immediate consumer decisions but also shapes long-term expectations and utility, leading to broader macroeconomic consequences.

Recent research in tourism economics literature provides fresh evidence that terrorism and related insecurity shocks have lasting macroeconomic consequences for tourism demand. Krajňák (2021) offers a systematic review covering two decades of empirical work and shows that terrorism consistently depresses tourist arrivals and receipts, with the magnitude and duration of losses depending on attack intensity and persistence. This persistence mirrors our “long-lasting fear” trajectory, where the perceived-safety index $\theta \left(t\right)$ recovers slowly and suppresses saving and investment. Complementing this, Ballis et al. (2019) document that geopolitical risk acts as a significant macro driver of international tourism, generating asymmetric responses across emerging economies; in our framework, such exogenous geopolitical shocks directly lower baseline $\overline{\theta }$ and reduces steady-state welfare. Furthermore, Harb and Bassil (2020) demonstrate that the negative terrorism-tourism relationship is moderated by immigration networks, which appear to restore trust and mitigate fear contagion. These findings motivate our behavioral mechanism in which social or informational capital cushion shocks to perceived security and shorten the recovery path of $\theta \left(t\right)$ .

Despite these insights, most existing studies remain empirical or descriptive, with limited formal modelling of the dynamic interaction between fear, consumption, investment, and economic growth. We address this gap by developing a continuous-time Ramsey-Cass-Koopmans (RCK) model that incorporates behavioural security perceptions into the household utility function. We model tourism as a distinct consumption category whose marginal utility is distorted by perceived safety, represented by a time-varying index $\theta \left(t\right)$ . A terrorism event induces a negative shock to $\theta \left(t\right)$ , capturing the erosion of public confidence. Our model allows this perception to recover over time, either endogenously (via psychological adjustment) or through exogenous policy interventions, such as public safety investments, risk communication campaigns, or law enforcement efforts.

We explore three representative scenarios: (i) a one-time shock with moderate recovery; (ii) a persistent shock with prolonged fear; and (iii) an intervention-based recovery driven by government policy. Through analytical derivations and numerical simulations, we show how different trajectories of perceived safety affect consumption allocation, capital accumulation, and long-term welfare. Our main findings are threefold. First, persistent terrorism-induced fear generates the largest long-run welfare losses by depressing both tourism and capital accumulation. Second, while government intervention can accelerate recovery, aggressive short-run stimulus to restore tourism may crowd out investment and trigger capital collapse if not properly timed. Third, optimal policy must balance between restoring confidence and maintaining investment incentive, avoiding excessive intertemporal shifts in consumption behaviour.

Our contribution lies in bridging behavioural macroeconomics and tourism economics by providing a formal theoretical model that captures the long-run economic implications of terrorism-induced fear. The findings have practical relevance for policymakers, particularly in designing post-crisis recovery strategies that align tourism promotion with sustainable economic growth.

Model

Set-up

In this section, we develop a Ramsey-Cass-Koopmans (RCK) model to analyse the dynamic effects of terrorism-induced fear on tourism demand and economic growth. We embed a behavioural security perception mechanism into a representative household’s intertemporal optimisation problem, and explore the role of government intervention in restoring confidence and stabilising long-run welfare.

Time is continuous, $t\in [0,\infty )$ . A representative infinitely-lived household derives utility from general consumption, denoted by $c\left(t\right)$ , and from tourism-related consumption, denoted by $l\left(t\right)$ . We assume the household faces security-related fear or anxiety that modifies the marginal utility of tourism. The instantaneous utility function is given by:

$U={\int }_0^{\infty }{e}^{-\rho t}\left[\ln \hspace{0.17em}c\left(t\right)+\theta \left(t\right)\ln \hspace{0.17em}l\left(t\right)\right]dt,$

(1)

where $\rho >0$ is the discount factor, and $\theta \left(t\right)\in \left[0,1\right]$ represents household’s time-varying perception of safety in the context of tourism. When $\theta \left(t\right)=0$ , fear is extreme and no utility is derived from tourism; when $\theta \left(t\right)=1$ , tourism is perceived as fully safe.

Household allocates total income to consumption, tourism, and savings. Capital accumulation evolves according to:

$\dot{k}\left(t\right)=f\left(k\left(t\right)\right)-c\left(t\right)-l\left(t\right)-\delta k\left(t\right),$

(2)

where $k\left(t\right)$ denotes capital per capita, $\delta >0$ is the depreciation rate.

We model terrorism as a negative psychological shock to tourism demand through the time-varying fear factor $\theta \left(t\right)$ . In the absence of government action, we assume:

$\theta \left(t\right)=\overline{\theta }-\varphi X\left(t\right),\varphi >0$

where $\overline{\theta }\in (0,1]$ represents baseline safety perception in the absence of terrorism; $X\left(t\right)$ stands for the terrorism intensity process (e.g., a Poisson shock), and $\varphi$ captures the marginal disutility of fear. A higher $X\left(t\right)$ reduces $\theta \left(t\right)$ , diminishing tourism utility and expenditure.

Firm’s production function is Cobb-Douglas: $f\left(k\right)=A{k\left(t\right)}^{\alpha }$ , where $A>0$ denotes total factor productivity (TFP), and $0<\alpha <1$ refers to output elasticity of capital. We introduce a policy function $G\left(t\right)$ representing public security investment to mitigate the effect of terrorism on tourism. Government spending improves public confidence, partially offsetting the fear shock. We assume:

$\theta \left(t\right)=\overline{\theta }-\varphi X\left(t\right)+\phi G\left(t\right),\phi >0$

where $\phi$ measures the effectiveness of government action in restoring perceived safety. In our setup, $G\left(t\right)$ acts as a confidence-restoring public good. Spending can be interpreted as law enforcement deployment, public communication, infrastructure investment, or tourist protection initiatives.

Household’s problem

We now solve the household’s optimization problem. The representative household chooses consumption $c\left(t\right)$ and tourism-related consumption $l\left(t\right)$ to maximize intertemporal utility subject to the capital accumulation constraint. Form the current-value Hamiltonian:

$\mathcal{H}=\ln \hspace{0.17em}c\left(t\right)+\theta \left(t\right)\ln \hspace{0.17em}l\left(t\right)+\lambda \left(t\right)\left[f\left(k\left(t\right)\right)-c\left(t\right)-l\left(t\right)-\delta k\left(t\right)\right],$

(3)

The first-order conditions (F.O.C) are:

$\frac{\partial \mathcal{H}}{\partial c}=\frac{1}{c\left(t\right)}-\lambda \left(t\right)=0\hspace{1em}⟹\hspace{1em}\lambda \left(t\right)=\frac{1}{c\left(t\right)}$

(4)

$\frac{\partial \mathcal{H}}{\partial l}=\frac{\theta \left(t\right)}{l\left(t\right)}-\lambda \left(t\right)=0\hspace{1em}⟹\hspace{1em}l\left(t\right)=\theta \left(t\right)·c\left(t\right)$

(5)

$\dot{\lambda }\left(t\right)=\rho \lambda \left(t\right)-\frac{\partial \mathcal{H}}{\partial k}=\lambda \left(t\right)\left(\rho +\delta -f^{\prime }\left(k\left(t\right)\right)\right)$

(6)

Substituting equation (4), the consumption growth equation becomes: ¹

$\frac{\dot{c}\left(t\right)}{c\left(t\right)}=f^{\prime }\left(k\left(t\right)\right)-\delta -\rho$

Steady-state

In the steady state, $\dot{c}\left(t\right)=\dot{k}=0$ and $\theta \left(t\right)=\overline{\theta }$ . The conditions simplify to:^2,3

$f^{\prime }\left(k^{*}\right)=\delta +\rho \hspace{1em}\Rightarrow \hspace{1em}{k}^{*}={\left(\frac{\alpha A}{\delta +\rho }\right)}^{1/\left(1-\alpha \right)}$

(7)

$c^{*}=\frac{f\left(k^{*}\right)-\delta k^{*}}{1+\overline{\theta }},l^{*}=\overline{\theta }·c^{*}$

(8)

The steady-state level of capital, consumption, and tourism are fully determined by fundamentals and the perceived security level $\overline{\theta }$ . Our model dynamics can be represented by a two-dimensional differential equation system in ( $k,c)$ space:

$\{\begin{array}{l}\dot{c}=c\left[f^{\prime }\left(k\right)-\delta -\rho \right]\\ \dot{k}=f\left(k\right)-\left(1+\theta \left(t\right)\right)c-\delta k\end{array}$

Figure 1 shows the phase diagram. The equation for blue dashed line is: $f\left(k\right)=\left(1+\theta \right)c+\delta k$ , which refers to $\dot{k}=0$ (i.e., capital no-change condition). That is, at any point along this curve, capital neither increases nor decreases. In the upper-left region, $\dot{k}<0$ (i.e., capital decreasing), while in the lower right region $\dot{k}>0$ (i.e., capital increasing). The equation for red dash-dot line is: $f^{\prime }\left(k\right)=\delta +\rho \Rightarrow k^{*}={\left(\frac{\alpha A}{\delta +\rho }\right)}^{\frac{1}{1-\alpha }}$ . It is a vertical line, representing the golden rule level of capital. In the left region, $\dot{c}<0$ (i.e., consumption decreasing), while in the right region $\dot{c}>0$ (i.e., consumption increasing). The intersection of the two curves is the steady-state equilibrium. We can see clearly that the unique saddle-path equilibrium converges to the steady state. Moreover, a positive security (e.g., a reduction in terrorism or an increase in government safety investment increases) $\theta \left(t\right)$ , reduces the required consumption share devoted to tourism, and encourages faster capital accumulation. Conversely, a persistent decline in $\theta \left(t\right)$ can trap the system in a low-growth regime with suppressed tourism demand and underinvestment.

Figure 1.

Phase diagram in ( $k,c)$ space under $\theta =0.5$ .

Simulation

We first simulate the dynamic adjustment of the security perception index $\theta \left(t\right)$ following a terrorism shock under three different scenarios. Figure 2 illustrates the simulated dynamic paths of $\theta \left(t\right)$ under three different post-terrorism scenarios. Specifically, we consider the following three trajectories of $\theta \left(t\right)$ : (i) one-time shock: temporary decline in security perception, modelled as an exponential decay function: $\theta \left(t\right)=\overline{\theta }-0.3e^{-0.2t}$ ; (ii) persistent shock: a long-lasting fear response where confidence recovers slowly: $\theta \left(t\right)=\overline{\theta }-0.3e^{-0.02t}$ ; and (iii) government intervention: public investment restores confidence more rapidly: $\theta \left(t\right)=\overline{\theta }-0.3e^{-0.4t}$ . In all cases, the shock initially reduces perceived safety from its baseline level $\overline{\theta }=0.5$ to $\theta =0.2$ , capturing a sudden loss in public confidence following an attack.

Figure 2.

Dynamic paths of security perception $\theta \left(t\right)$ under terror shock.

The one-time shock scenario (solid yellow curve) models an exponentially decaying fear response, where public confidence gradually returns in the absence of follow-up attacks. This case captures a typical psychological recovery from a single event. The persistent shock scenario (dashed orange curve) assumes a much slower recovery process, mimicking situations where repeated minor threats, prolonged media exposure, or structural fear inhibit the restoration of perceived safety. In this case, $\theta \left(t\right)$ remains significantly below its pre-shock level for an extended period, implying a sustained drag on tourism demand and economic performance. In contrast, the government intervention scenario (dash-dotted red curve) introduces a policy-induced recovery mechanism, where public spending and confidence-building efforts accelerate the rebound of $\theta \left(t\right)$ . This results in a steeper and earlier convergence back to the baseline. Figure 2 highlights the central role of behavioural fear dynamics and the potential for timely public policy to mitigate long-term macroeconomic damage from terrorist events.

Next, to assess the quantitative implications of terrorism-induced fear on tourism and macroeconomic dynamics, we again simulate the model under three distinct scenarios for the evolution of $\theta \left(t\right)$ . All scenarios begin with an identical magnitude of fear shock (30% drop), but differ in recovery speed. $\overline{\theta }$ is normalized to 0.5 across simulations. We numerically solve the nonlinear system consisting of:

$\frac{\dot{c}\left(t\right)}{c\left(t\right)}=f^{\prime }\left(k\left(t\right)\right)-\delta -\rho ,\dot{k}\left(t\right)=f\left(k\left(t\right)\right)-\left(1+\theta \left(t\right)\right)c\left(t\right)-\delta k\left(t\right)$

Using a forward-shooting method over a 50-year horizon, initializing from $k_0=2$ and $c_0=0.5$ , the resulting trajectories $k\left(t\right)$ and $c\left(t\right)$ exhibit distinct dynamics.

Figure 3 shows the dynamic evolution of capital per capita $k\left(t\right)$ over a 50-period horizon under three distinct security scenarios corresponding to different trajectories of $\theta \left(t\right)$ . In the one-time shock scenario (orange curve), confidence gradually returns to normal, allowing the economy to maintain a positive savings rate. As a result, capital accumulation continues, albeit at a slower pace than in the steady-state path without shocks. In contrast, the persistent shock scenario (yellow curve) features a prolonged fear environment where $\theta \left(t\right)$ recovers very slowly. This sustained anxiety depresses consumption-adjusted saving, but the economy still manages to accumulate capital over time, albeit more sluggishly. The most striking outcomes arises in the government intervention scenario (red curve), where confidence is restored rapidly due to public safety policies. Surprisingly, capital collapses after a temporary rise. This counterintuitive result stems from the fact that rebound in $\theta \left(t\right)$ increases the household’s effective marginal utility from tourism consumption so sharply that optimal consumption surges, crowding out investment. Once capital drops below a certain threshold, depreciation outweighs output, triggering a downward spiral and collapse toward a capital-poor steady state. This demonstrates a potential policy pitfall: if demand is overstimulated before productive capacity has recovered, the economy may overshoot its short-term consumption potential, leading to unsustainable trajectories.

Figure 3.

Capital dynamics $k\left(t\right)$ under different security scenarios.

Figure 4 illustrates the simulated time paths of consumption per capita $c\left(t\right)$ over a 50-period horizon, under three distinct trajectories of $\theta \left(t\right)$ . The figure reveals sharp differences in consumption dynamics depending on the speed and persistence of the security shock. In one-time shock scenario (yellow curve), consumption initially rises as investment gradually recovers, but then declines in later periods due to the compounding effects of diminished capital accumulation. The persistent shock scenario (orange curve) results in an even more pronounced long-run decline in $c\left(t\right)$ , reflecting sluggish recovery of public confidence and a sustained drag on both savings and output. By contrast, the government intervention scenario (red curve) produces a dramatically different outcome: consumption increases rapidly as confidence is restored and the perceived utility from tourism rebounds. However, the rise in consumption becomes unsustainable as it outpaces capital growth, eventually hitting an imposed cap in the simulation (reflecting resource constraints). This pattern illustrates how aggressive stimulus to tourism-related utility can shift preferences toward immediate consumption, which—if not matched by productive capacity—risks depleting capital and destabilizing the long-run growth path.

Figure 4.

Consumption dynamics $c\left(t\right)$ under different security scenarios.

Overall, Figure 4 emphasizes that faster recovery of perceived safety boosts short-run consumption, but the macroeconomic effects depend critically on the interplay between behavioural responses, capital formation, and sustainability constraints. Policymakers must calibrate interventions not only to restore confidence, but to avoid excessive intertemporal shifts in household behaviour that could undermine long-term economic resilience.

Welfare comparison and losses

Figure 5 shows the welfare losses associated with three different trajectories of $\theta \left(t\right)$ , following a simulated terrorism shock. We first numerically solve the RCK model under each $\theta \left(t\right)$ scenario, as described earlier, and compute the lifetime utility of the representative household: $W={\int }_0^T{e}^{-\rho t}\left[\ln \hspace{0.17em}c\left(t\right)+\theta \left(t\right)\ln \hspace{0.17em}l\left(t\right)\right]dt$ , where $T=50$ is the planning horizon. Welfare loss are then measured as the percentage reduction in total utility relative to the best-performing scenario—government intervention—where confidence recovers rapidly. ⁴

Figure 5.

Welfare loss under different $\theta \left(t\right)$ scenarios.

Figure 5 reveals that the persistent shock (orange bar) yields the highest welfare loss, amounting to nearly 25 utility units below the best-case outcome. This reflects the prolonged period of reduced perceived safety, which depresses both direct utility from tourism and investment-driven growth. The one-time shock (blue bar) performs moderately better, but still exhibits substantial welfare loss due to the slower economic rebound in the absence of policy intervention. As expected, the government intervention results in no welfare loss by construction, since it serves as the benchmark.

In summary, the figure underscores the macroeconomic cost of fear persistence, and highlights the powerful welfare-enhancing role of timely and effective government action. While even a temporary confidence shock can reduce long-term welfare, persistent uncertainty exerts particularly damaging and irreversible effects.

The dynamic of our model is consistent with stylized facts from major terrorism incidents. In the aftermath of the 9/11 attacks in the United States, international tourist arrivals fell sharply and the recovery of traveller confidence was protracted, with flows taking several years to return to pre-attack levels. This trajectory mirrors our persistent fear scenario, where slow restoration of perceived safety suppresses saving, capital accumulation, and long-run welfare. By contrast, following the 2015 Bataclan terrorist attack in Paris, the French government deployed strong security measures and communication campaigns to reassure visitors. Although arrivals initially dropped steeply, confidence and tourism flows recovered relatively quickly compared with the U.S. case. This experience resembles our policy-driven recovery scenario, in which public intervention accelerates the rebound in perceived safety but also requires careful calibration to avoid crowding out investment. Together, these stylized facts reinforce the relevance of the behavioural fear channel in our framework and demonstrate that the trade-offs highlighted in the simulations are not purely theoretical but also observed in real-world post-terrorism tourism dynamics.

Threshold effects of government intervention

To further investigate the dynamics of government intervention, we extend the baseline simulations by varying the intensity of public safety investment, captured by the parameter $g$ , which accelerates the recovery of perceived safety $\theta \left(t\right)$ . A higher value of $g$ implies stronger or faster government efforts to restore public confidence through measures such as security investment, marketing campaigns, or risk communication. Figures 6–8 plot the trajectories of capital $k\left(t\right)$ , consumption $c\left(t\right)$ , and safety perception $\theta \left(t\right)$ under alternative intervention intensities. The results highlight a clear non-linear trade-off. For low or moderate levels of intervention, faster confidence recovery raises consumption while still allowing capital accumulation, consistent with the stabilizing mechanism described in sections Household's problem to Simulation. However, when intervention becomes aggressive, the rapid rebound in $\theta \left(t\right)$ raises the marginal utility of tourism consumption so strongly that households reallocate disproportionately toward immediate consumption. As shown in Figure 6, this shift reduces saving and slows capital accumulation, leaving the economy with a lower long-run capital stock compared to more gradual recovery paths. The consumption trajectories in Figure 7 confirm that while households enjoy higher short-run utility, this is achieved at the expense of long-term productivity capacity.

Figure 6.

Capital dynamics under varying government intervention intensity $g$ .

Figure 7.

Consumption dynamics under varying government intervention intensity $g$ .

Figure 8.

Perceived safety recovery under varying government intervention intensity $g$ .

Importantly, the extended simulations suggest that under the baseline calibration, no outright collapse of $k\left(t\right)$ or $c\left(t\right)$ occurs within plausible ranges of government intervention. Instead, the evidence points to a threshold effect: beyond a certain intensity of intervention, additional spending ceases to improve long-run welfare and instead crowds out investment. This finding underscores the policy pitfall identified earlier in section 2.4—that overstimulating demand before capital has recovered risks undermining macroeconomic stability. For policymakers, the implication is that government action should be carefully calibrated. In particular, insufficient intervention prolongs fear and welfare loss, while overly aggressive intervention undermines the very growth foundations required for sustainable recovery.

Managerial and policy implications

Our results regards the government intervention scenario carry clear implications for both managers in the tourism sector and policymakers. From a managerial perspective, rapid restoration of traveller confidence—though attractive in the short run—can generate unintended macroeconomic vulnerabilities if demand surges faster than productive capacity. Destination managers and firms in hospitality, transport, and culture services should therefore align recovery campaigns and promotional efforts with the gradual rebuilding of infrastructure, workforce capacity, and investment. This implies that recovery strategies should prioritize resilience-building measures such as skill training, diversification of tourism products, and phased promotional activities, rather than relying solely on short-term marketing or subsidies.

For policymakers, our model highlights a delicate trade-off in post-terrorism recovery design. Timely public spending on safety, infrastructure, and communication can accelerate confidence restoration, but if implemented too aggressively, it risks crowding out investment and destabilizing long-run growth. Calibrated, time-phased interventions are therefore essential. Recent OECD reports emphasize the importance of staging fiscal and promotional measures so as to balance immediate relief with investment sustainability (OECD, 2020; OECD, 2022). Similarly, policy analyses stress that stimulus measures are most effective in less resilient economies, while more resilient destinations require more nuanced strategies (Okafor et al., 2022). The broader implication is that public recovery efforts must be carefully tailored to local economic structures, with governments deploying confidence-building policies in parallel with long-term investment frameworks.

Taken together, our results suggest that successful recovery strategies require integrated coordination between government agencies and private sector actors. Tourism recovery is not only about restoring confidence but also about safeguarding macroeconomic resilience. As the OECD (2022) underscores, post-crisis periods offer a window of opportunity to embed sustainability, inclusiveness, and resilience into tourism policy frameworks. Our theoretical findings therefore complement applied policy advice by demonstrating, in a formal setting, the potential risks of overstimulation and benefits of calibrated, staged interventions that ensure both short-term recovery and long-term growth.

Conclusion

This paper develops a dynamic macroeconomic framework to analyse the long-run effects of tourism-induced fear on tourism and growth. By embedding a time-varying perception of security into a RCK model, we formally incorporate the behavioural distortions that arise when fear alters the marginal utility of tourism consumption. Our model captures how tourism shocks, though non-economic in origin, can trigger substantial and persistent economic disruptions through changes in consumption, savings, and capital accumulation.

Our analysis shows that the fear channel operates as a behavioural wedge that depresses tourism demand, reduces effective saving rates, and impedes long-term capital formation. This distortion, if left unaddressed, can push the economy toward a low-growth trajectory characterized by underinvestment and declining welfare. Our model also reveals that the recovery path of perceived safety—whether fast, slow, or policy-induced—critically determines the severity and persistence of macroeconomic losses.

Simulation results across three fear trajectories highlight the nonlinear and often counterintuitive dynamics of post-terror recovery. A one-time shock with natural fear decay leads to moderate welfare loss, while persistent fear results in the most severe long-run consequences. Interestingly, our model cautions against overly aggressive government intervention aimed at rapidly restoring tourism demand. If public policies stimulate tourism-related consumption before productive capacity has recovered, the resulting shift away from savings can induce capital collapse, undermining long-run growth and stability. Our findings thus call for calibrated, rather than purely reactive, post-crisis tourism and security policies—ones that rebuild confidence while preserving macroeconomic resilience.

Future research can extend the model in several directions. One promising avenue is to endogenize government behaviour, allowing public investment in security to be determined optimally in response to shock size and economic conditions. Another is to incorporate international spillovers—where tourism in one country affects tourism demand in neighbouring regions—within a multi-country general equilibrium framework.

Ultimately, our work underscores that while terrorism is a physical act of violence, its most enduring economic damage often arises from fear—and that managing this fear is as much an economic challenge as a psychological one.