Fear of Violence during Armed Conflict: Social Roles and Responsibilities as Determinants of Fear

Abstract

This article investigates the prevalence and determinants of fear as a consequence of living through armed conflict. We use survey data from Nepal during the armed conflict (1996–2006) to examine how trauma, sex and gender, age, marriage, and household size affect fear of violence. We also disaggregate types of worry, and find substantial variance on whether respondents were more concerned about livelihood consequences of conflict than physical danger. We supplement quantitative analyses with discussion of in-depth interviews from the study area on these same topics. Overall, our results highlight the enduring impact of gender roles in Nepal and that conflict might disproportionately affect those who are already vulnerable and have greater social responsibilities. This article provides a unique comparison between fear of violence during armed conflict in a low-income country to the fear of crime literature based in high-income countries.

1. INTRODUCTION

Quantitative research on the consequences of armed conflict on civilians has largely focused on human casualties from violence and specific morbidities, such as injuries, infectious diseases, malnutrition, and mental health disorders like post-traumatic stress disorder (PTSD) and depression (Bolton and Betancourt 2004; Garfield 2007; Ghobarah, Huth, and Russett et al. 2003; Krug et al. 2002; Levy and Sidel 2000; Mack 2005; Murray et al. 2002; Pedersen 2002; Sonis et al. 2009; Thapa and Hauff 2005). Beyond the direct consequences of living through an armed conflict – such as physical injuries, disruptions to daily life and routine mobility, and poor living conditions (e.g., limited and unsafe water and food supplies and poor access to health care), such an environment leads to heightened insecurity and unpredictability in daily life. Recent research on trauma and mental health suggests that chronic feelings of insecurity and stress might be one unexplored mechanism through which conflict affects well-being and health outcomes (Green et al. 2003; Momartin et al. 2006; Steel et al. 2006; Steel et al. 2009; Thapa and Hauff 2005). Despite this suggestion, these indirect mechanisms connecting armed conflict with well-being and health outcomes have received little attention in the literature (Miller and Rasmussen 2010; Panter-Brick 2010; Pedersen 2002; Tol et al. 2010). This key gap in the literature leaves us with an incomplete understanding of how the broader context of armed conflict, including the concomitant stress and fear, might impact individuals and families immediately and in the long-term.

In this paper we seek to improve our understanding of one of the indirect, but possibly influential, mechanisms that might connect the experience of armed conflict to health and wellbeing outcomes—stress and insecurity. Given that this is a broad concept, in this study we operationalize stress and insecurity more specifically as fear of violence. We examine how individual-level characteristics, particularly related to social roles and responsibilities, influence variance in levels of fear of violence during armed conflict.

We conceptualize fear of violence as a negative emotional response to a situation that is perceived to be threatening. In the case of armed conflicts, which can last for long periods of time, fear of violence manifests as chronic feelings of insecurity and threat. Fear of violence is an important result in its own right as it affects quality of life in multiple ways. As a form of chronic stress, fear of violence might also be a key mechanism that affects other mental and physical health outcomes during and post-conflict, including PTSD, depression, and anxiety (Christopher 2004; Gunnar and Quevedo 2007; Ferry et al. 2008; Miller and Rasmussen 2010). In addition, fear of violence can influence important health-related behaviors such as movement within the community, the development of supportive social relationships, utilization of health services, marriage, childbearing, and migration (Agadjanian and Prata 2002, 2001; Caldwell 2004; Cullen and Levitt 1999; Das-Gupta and Li 1999; De Smedt 1998; Lester 1993; Lindstorm and Berhanu 1999; Rountree and Land 1996a; Saxena, Kulczucki and Jurdi 2004). On the other hand, fear is also associated with worry, which research shows can have negative consequences as well as positive consequences on motivating productive or safety-seeking behaviors and as an emotional buffer (Sweeny and Dooley 2017).

Despite the serious short- and long-term consequences of chronic fear of violence during armed conflict, there is limited attention in the literature on the question of who is likely to experience fear and why. An important exception is several recent ethnographic studies of fear during conflicts in Nepal, Guatemala, and Northern Ireland (Ferry et al. 2008; Green 1995; Lysaght 2005; Pettigrew 2003; Pettigrew and Adhikari 2009). For example, in an ethnographic study of fear during the recent Nepali conflict, Pettigrew and Adhikari identify variations in fear related to, “personality, the impact of particular experiences and life situations, coping skills and available resources including degrees of support.” (Pettigrew and Adhikari 2009:412). However, we are not aware of any systematic quantitative studies that address the social, cultural, and economic factors that influence the development of fear of violence during armed conflict.

Another area of study—fear of crime—hosts a substantial body of theoretical and empirical literature that provides some important insights and can help guide the study of fear of violence during conflict. Fear of crime research has linked higher levels of fear to perceived risk and vulnerability, poverty, physical vulnerability, sex and gender, age, and race and ethnicity (Ferraro 1996; Franklin and Franklin 2009; Hale 1996; Killias 1990; Killias and Clerici 2000; Madriz 1997; Pantazis 2000; Rader, Cossman and Porter 2012; Riger at el., 1978; Snedker 2012; Warr and Stafford 1983). Moreover, fear of crime can include both fear for oneself and fear of crime for others known as altruistic fear (Warr and Ellison 2000).

In addition to personal vulnerability, fear of crime research also addresses the influence of community context. Neighborhood conditions such as perceived disorder and incivilities, lack of social integration, and lack of community social capital are associated with greater levels of fear (Ferguson and Mindel 2007; Franklin, Franklin and Fearn 2008; Ross and Jang 2000; Skogan and Maxfield 1981; Wyant 2008). Notions of ‘ecological vulnerability’ are linked with higher levels of fear, especially for women (Hale 1996). However, the vast majority of these studies are based in the United States and Western Europe in the context of generalized political peace¹ and mostly in urban contexts. There exists little understanding of how the development of fear might be different between high-income and low income settings or universal in both, during armed conflicts that are characterized by long-term and widespread violence, and in populations not related to the European diaspora.

Borrowing insights from the fear of crime literature, we assess the applicability of existing theories to understanding fear of violence during armed conflict in Nepal. Do expected findings from fear of crime research produce similar results in terms of fear of violence in a context of armed conflict? What additional insights can we gain about fear of crime in general, by applying theories to an entirely new context? Our empirical analyses are based primarily on statistical analysis of detailed survey data and supplemented with in-depth interview data that aids in our interpretation of statistical results.

In the next section, we describe the context of this study, the Chitwan Valley in rural Nepal, and the conflict that occurred there between 1996–2006. After that, we present a theoretical discussion of the ways in which social roles and responsibilities might influence the development of fear of violence during armed conflict. In that section, we draw heavily on the fear of crime literature. We discuss how these insights might be more and less applicable in the context of armed conflict and how they might apply to the specific context of rural Nepal.

2. CONTEXT AND SETTING

Our case study is based on the decade-long armed conflict between the Communist Party of Nepal (Maoist) and the government security forces in Nepal. The study site is the western Chitwan Valley of south-central Nepal. The valley is flat, fertile, and dominated by agriculture. The administrative district of Chitwan borders India and is about 100 miles from Kathmandu. There is one large city, Narayanghat, and the rest of Chitwan’s population, like much of Nepal, lives in small rural villages.

Armed conflict officially began in 1996 when the Communist Party of Nepal (Maoist) made a formal declaration of “People’s War” following a relatively unsuccessful political campaign. The aim of their movement was to unseat the constitutional monarchy and install a democratic republic. They charged the government with poor administration, corruption, unfair taxation, and neglect of poor rural areas of the country.

The earlier stages of the conflict were contained primarily in several mid-western districts, the most remote and under developed part of the country, and aimed at damage to government installations and communication infrastructure, capturing weapons, and threatening government security forces. From mid-2000 however, the Maoists progressively expanded their campaign nationwide. In January 2001, the Nepalese government responded by creating a special armed police force to fight the Maoists. After that, the government generally maintained control of cities and large towns and the Maoists controlled a majority of the rugged countryside of Nepal, where communication and transportation are difficult. By 2001, they were operating in 68 of Nepal’s 75 administrative districts including Chitwan where this study is based. After a series of feigned negotiations, ceasefires, and repeated talks, in June 2006 serious peace talks commenced and on November 21, 2006 the government and Maoists signed a comprehensive peace agreement, ending the armed conflict.

Because this armed conflict was staged mainly using guerrilla tactics, there was generally no ‘frontline’, it was largely unknown where fighting would break out, and civilians were often unintentionally caught up in firefights and bomb blasts. In addition, both Maoists and government forces intentionally used civilians for political purposes. Reported violent acts by the Maoists and Nepalese government security forces against civilians include torture, extrajudicial killings, bombings, gun fights, abductions, forced conscription, billeting, taxing, extortion, and general strikes (Hutt 2004; Pettigrew 2004). Three ceasefires were called and subsequently broken. The government called a State of Emergency and instituted martial law twice, in 2001 and 2005. From 1996 until the end of 2006, there were an estimated 17,000 fatalities due to the conflict, and responsibility for these deaths lies with both the Maoists and the Government Security Forces (Informal Sector Service Center 2006).

Throughout the conflict, the Chitwan Valley, from which the data for this study were collected, experienced a moderate amount of violence. Between 1996 and April 2006, Chitwan experienced 194 conflict related fatalities, slightly higher than the average number of fatalities of all districts (Informal Sector Service Center 2006). During the conflict, Chitwan experienced a number of bomb blasts, the great majority in 2003 and 2004, the largest of which injured or killed 17 people². There was one major gun battle in Chitwan between Maoists and security forces in June 2005 that resulted in 34 civilian fatalities. Along with these visible and countable disturbances, the people of Chitwan were regularly subjected to house raids, beatings, intimidation, arrests without warrant, torture and extra-judicial killings, taxes, billeting, conscription (by both Maoists and the government), curfews, and general strikes.

3. THEORETICAL BACKGROUND

In seeking to understand variance in the development of fear of violence during armed conflict, it is important to consider the macro-level exposure to conflict. In any conflict, some events such as bombs and gun battles are geographically and temporally specific, they occur in particular areas at particular times. Targeted violence, such as assassination and assault, is even more specific, affecting some individuals and not others. Thus the first clue towards understanding variance in fear of violence is to specify exposure to events. To this end, some studies in political science and demography use an event-centered approach by specifying which individuals experienced different kinds of violent and political events and using that to predict outcomes (Bond et al. 1997; Davenport and Stam 2009; Jenkins and Bond 2001; Schrodt and Gerner 2004; Williams et al. 2012).

We approach this study with the additional proposition that even among those exposed to the same violent or political events, there will be important individual-level variation in the development of fear. Indeed, quantitative research on fear of crime already identifies several social demographic factors and contextual conditions that influence variation in fear of crime. Factors that contribute to the emotional development of fear include underlying vulnerabilities – sex and gender, age, and income (Killias and Clerici, 2000; Pantazis, 2000) – and social roles and responsibilities for one’s own and others’ welfare (Jackson 2004). In fact, studies of altruistic fear explicitly link levels of fear to gendered social roles (Warr and Ellison 2000; Rader 2010; Snedker 2006).

In the context of armed conflict, we hypothesize that similar factors will impact fear of violence and that social roles and responsibilities will play a key role in the development of fear. As in all social patterns, the exact context (rural Nepal or urban Pittsburgh) will determine the meaning and application of social roles and responsibilities, and thus the exact connection between each social-demographic factor and fear of violence.

In considering variations of fear of violence during armed conflict, our second theoretical proposition is that there are a variety of consequences to armed conflict, in addition to insecurity and danger. Economic problems include disruptions in the ability to perform daily social and livelihood activities. Commodity prices and tax rates often increase, and looting and destruction of homes, farms, and other assets is common (Bundervoet and Verwimp 2005; Collier 1999; Justino 2006; Mack 2005). Social and political life can also be disrupted, with changes in school and health center access, patterns of marriage and fertility, and social relationships and leadership hierarchies (Agadjanian and Prata 2002; Caldwell 2004; Heuveline and Poch 2007; Justino 2006; Lindstrom and Berhanu 1999; Mack 2005; Shemyakina 2006; Williams et al. 2012). These indirect consequences of armed conflict can also contribute to the development of fears for economic and social security. In other words, individuals might develop fear in response to physical insecurity, but also in response to social and economic insecurity.

In the remainder of this section, and paper, we focus on several socio-demographic factors that might influence fear of violence during armed conflict- sex, marital status, household size, and age. Drawing from the fear of crime literature, we focus on how these sociodemographic factors delineate social roles and responsibilities, and it is these roles and responsibilities for particular aspects of personal and family life that help us to understand why some people have greater fear than others for different consequences of conflict. Although not related to social roles and responsibilities, we also include a short discussion of how direct traumatic experience influences fear. This draws on the theoretically and empirically strong relationship between trauma and fear. Our analysis uses the event-centered approach by measuring the actual conflict-related events to which an individual was exposed instead of assuming homogeneity of experience of conflict.

3.1 Trauma

Sociological studies of fear of crime show that higher levels of fear persist among people with recent history of personal victimization (Lee 1982; Roundtree, and Land 1996a; Skogan and Maxfield 1981; Stafford and Galle 1984;). Psychological research finds a similar relationship, where experience of traumatic events increases an individual’s fear of violence (Frijda 1993; Terheggen, Stroebe and Kleber 2001). Indeed, so strong is the likely relationship between trauma and fear, that we would be remiss in excluding it from our study. Thus we predict that experience of traumatic events will increase fears of violence during armed conflict (H1).

3.2 Sex and Gender Roles

A host of socio-demographic factors are related to fear of crime but sex is the best documented, with consistent findings that women report greater fear of crime than men (Ferraro 1995; Fisher and May 2009; Fox et al. 2009; Franklin and Franklin 2009; Snedker 2015). Existing evidence suggests that gender is associated with fear of violence in several ways. First, depending on social context, research in social-psychology has shown that men and women can cognitively process risk and fear of victimization differently (Das-Gupta and Li 1999; Rountree and Land 1996b; Smith and Torstensson 1997; Warr 1984). Specifically, evidence shows that given similar perceptions of their expected risk of victimization, women have higher levels of fear than men. This could be for several reasons, including differences in self-efficacy or that women, particularly in patriarchal societies such as Nepal (Bennett 1983; Acharya and Bennett 1981), are socialized to believe that they are more vulnerable, less able to control their situation, and less able to cope with psychological stress induced by violence (Das-Gupta and Li 1999; Gabriel and Greve 2003; Skogan and Maxfield 1981; Sutton and Farrall 2005; Stanko 1990; Warr 1984). Not only are there gender differences in socialization about self-efficacy, but there are also differences in the actual agency and autonomy of women; in many places women’s ability to make decisions for themselves and family members is restricted, which could lead to greater fear. Further, it may be that women are more aware of danger and more sensitive to risk as well as more perceptive to the signals of danger (Smith and Torstensson 1997; Stanko 1990). On the other hand, for men, gender role socialization and stereotypes about masculinity may suppress men’s willingness to report fear of crime (Gilchrist et al. 1998) as a sign of weakness (Goodey 1997). Thus, a significant part of the explanation for sex differences in fear of crime is related to gender roles and expectations.

Gender roles are also important in fear for others (altruistic or vicarious fear). Women are often responsible not just for their own well-being, but also for others in the family, including those who are particularly vulnerable, such as children and the elderly. Women are often socialized to express more fear in general and specifically for their children (parental fear), while men are socialized to express concern for those whom they feel a duty to protect, often their wives (spousal fear) and children (Snedker 2006; Rader 2010; Stanko 1990; Warr and Ellison 2000).

We expect these patterns to be similar in the case of armed conflict, where real and serious threats exist to both caretakers and protectors. Of course gendered differences in social roles, and subsequently in fear, are context dependent, but should be similar in the case of Nepal. In places where women are responsible for the health and general well-being of themselves and multiple family members, such as in Nepal, we predict that the frequency and amount of fear of violence they experience will be greater than that for men. On the other hand, men are often responsible for the financial well-being of the family, through farming, work, or business ownership. Just as armed conflict can threaten the safety of individuals and their family members, it can also threaten the financial well-being of the family by restricting the ability to get to the farm, to go to work, or direct threats on businesses. Thus, we expect women to experience greater fear than men. Further, we expect men to comparatively worry more about the financial consequences of conflict (such as disruptions in work, farming, loss of jobs, or loss of business) than women, and women to comparatively worry more about danger than financial consequences (H2).

3.3 Age

In fear of crime research, many studies report that age is a significant predictor of fear of crime; although the direction of the effect is debated. On one side of the debate, researchers claim that the elderly are more fearful (Baumer 1978; Hindelang, Gottfredson & Garofalo 1978; Pantazis 2000). As people age, they report greater fear and concern with crime due to their greater physical and social vulnerability (Bachman 1992). However, there might also be a gender dimension, as a disproportionate number of women are among the older population (Baumer 1978; Pain 1997) and Franklin and Franklin (2009) suggest that aging reduces fear of crime for women but not men. Alternatively, other researchers assert that younger respondents are more fearful or that there is a curvilinear relationship (Ferraro 1995; Garofalo and Laub 1978; LaGrange and Ferraro 1989). It might also be possible that perceived risk plays a role, but LaGrange and Ferraro (1989) report neither of their measures of perceived risk are strongly correlated with age.

Complicating the issue even more, some studies argue that the effect of age is somewhat less important than previously supposed and might in fact be overestimated (Clemente and Kleiman 1977; LaGrange and Ferraro 1989). Taking into consideration other variables, such as gender and income, and close attention to context might help to explain the supposed relationship between age and fear of crime (Ferraro and LaGrange 1992; Pain 1997).

This relatively extensive research leads us to no firm conclusions, but many ideas for how age and fear of violence might be related in the context of armed conflict. We might expect that older people will experience more fear due to vulnerability and comparatively greater perception of risk. However, taking into consideration the context of armed conflict (as suggested by Pain 1997), we note that during many conflicts younger people are specifically targeted for military conscription, other types of personal violence, and extortion (Thapa and Sijapati 2003; Wilson 2000). For example, during the armed conflict in Nepal, unmarried youths were targeted for forced and voluntary conscription by the Maoists and the Government security forces. Thus in addition to random experiences of violence (such as getting caught in a bomb blast or gun battle while in public areas), there are more types of violence to which younger adults are exposed, compared to their older counterparts. Moreover, younger adults are generally more mobile within the community, tasked with non-family work, going to the market, and other chores outside the home. This again exposes them to violence to a greater extent than their older counterparts. While there are theoretical reasons to predict a negative influence of age on fear of violence and other theoretical reasons to predict the opposite, we believe that the actual risk of violence during armed conflict was so much greater for young adults as to have a stronger influence on fear. Therefore, we expect younger adults to have greater fear of violence than older adults (H3).

3.4 Marriage and Household Size

In a similar manner, marital status and characteristics of the household may confer specific social roles and responsibilities on the individual and consequently influence the development of fear and type of fear experienced. Being married, associates adults with a wider social network of individuals to whom they are responsible. Married adults are more likely to have children, and as in the case of Nepal, often live with elderly or widowed parents for whom they also provide care. Furthermore, marriage is often associated with a transition to adulthood with new roles and expectations, particularly for the welfare of other family members. In a qualitative study of married men and fear or crime, marriage altered men’s fear of crime as they moved from personal fear to vicarious fear for their spouse (Rader 2010). Thus, from the perspective of social roles and responsibilities, we might expect marriage to increase fear of violence.

Alternately, there is a general perception that marriage increases the social support available to an individual. Evidence shows that social support is an important factor mitigating individuals’ experience of violence and other macro-level disasters (Carr et al. 1997; Cleary and Houts 1984; Creamer et al. 1993; Galea et al. 2002; Hobfoll, Tracy and Galea 2006; Norris et al. 2002; Norris and Kaniasty 1996; Tracy et al. 2008). Thus, from a social support perspective, we might expect marriage to decrease fear of violence, exactly the opposite prediction of the social roles and responsibilities perspective. Of course, it is also possible that both mechanisms could function at the same time. In this case, we must turn to the context of a particular study in order to adjudicate to what extent marriage might provide social support and if this support might be more or less consequential than the increased social responsibilities.

In Nepal, people most often marry at relatively young ages, quickly start childbearing, and live in extended family situations. As such, the responsibility to care for newborn and other family members is generally quite high. At the same time, gender roles within marriage are often very segregated, with women responsible for a set of tasks (childcare, cooking, home care, elderly care) that has little overlap with the set of tasks to which men are responsible (farm labor). Furthermore, relationships with co-resident in-laws are often difficult and complex. Indeed, it is relatively common for married women to take extended visits to their natal family to seek social support that they might not receive in their married household. In this setting, while marriage might bring increased social support and increased responsibility, we argue that it adds more responsibility than support. Thus, we predict that married individuals will report more fear than unmarried, and this effect will be stronger for women (H4).

We expect household size to have a positive effect on fear of violence similar to that of marriage. Living in a larger households means that adults are responsible for the financial and general well-being of more people. Larger households often encompass complex families, including elderly parents who might need extra care and are increasingly vulnerable. Thus we predict that individuals living in larger households will report more fear than those living in smaller households (H5).

4. METHODS AND MATERIALS

Data for this study come from a survey undertaken in the Chitwan Valley in 2009 designed to measure fear of violence, traumatic experiences, and mental health outcomes³. The survey was administered to a systematically selected sample of 403 individuals living in the western Chitwan Valley. Using the Chitwan Valley Family Study (CVFS)⁴ sampling frame, respondents were chosen using a two stage sampling procedure. First, the study area was divided into three distinct strata based on distance from Narayanghat, the one urban area in the northeast corner of the study area. Each stratum is progressively more rural with lower population density. A sample of four to six neighborhoods, each consisting of between 10 and 15 households, was randomly selected from each stratum⁵. In addition, based on a local psychiatric doctor’s recommendation, we purposively choose two additional neighborhoods from poor-slum areas that are likely to have high a prevalence of psychological and emotional distress⁶. Once a neighborhood was selected, all individuals age 18 and above residing in households (eating and sleeping most of the time in the last month) within the selected neighborhoods were interviewed. This sampling procedure resulted in 403 individuals being selected in a statistically representative manner that closely resembles the CVFS sample. For this study we used 401 people for whom full information was available on our measures of interest. These people were interviewed in face-to-face interviews in Nepali language, resulting in a 91 percent response rate.

In order to examine the representativeness of our sample, we compare it to the representative sample taken by the Chitwan Valley Family Study (CVFS), a much larger sample of 5,271 people in the same study area. Table 1 displays the demographic composition of the sample used in this study (fear of violence sample) and in the CVFS sample. As shown in Table 1, our study sample closely matches the large scale CVFS sample, which increases our confidence in the representativeness of our sample. In both surveys, there are more women (about 59 %) than men, primarily because of high male out-migration for employment and schooling. The ethnic distribution closely resembles the CVFS sample. The mean age is 37. Marital status in our sample shows the universality of marriage; people in the ‘Other’ category mostly include those who were widowed, with very few divorcees. Households are generally. large, with an average of about six people and include extended family members, often elderly parents.

Table 1.

Comparison between fear of violence sample and CVFS sample

	Fear of violence Sample (used in analysis)	CVFS Sample
Sex
Male	41%	40%
Female	59%	60%
Age
Mean age	37	38
Range	18–90	15–88
Marital Status
Never married	13%	13%
Married	83%	82%
Divorced, separated or widowed	4%	5%
Household size
Mean	6.11	5.46
Ethnicity
Brahmin-Chhetri and Newar	47%	48%
Dalit	20%	17%
Hill Janjati	13%	13%
Terai Janjati	20%	17%
Education
Mean # of years completed	5	7
At least 3 years completed	37%	20%
Range (in years)a	0–16	0–16

^a16 years of education completed refers to MA or greater level completion

In addition to the survey data we draw from 25 semi-structured interviews with residents of Chitwan. Respondents were from the same area as the survey and were purposively chosen to broadly mirror the age, sex, occupation, and education distribution of the area’s population. Interviews were undertaken in Nepali language with trained Nepali interviewers who worked closely with two of the authors of this study (Ghimire and Williams) to design appropriate broad topical guidelines for the interviews. We use these in-depth interviews to assist in our understanding and interpretation of statistical results and to provide nuance and further detail to the topics we present in this paper.

4.1 Outcome measures

The outcomes of interest in this study include three measures of fear of violence and worries about other consequences of conflict. With reference to the Maoist conflict of the previous decade, respondents were first asked, “What problem were you most worried about during the conflict?” Answer choices included: physical danger (such as getting killed, injured, beaten, abducted, or tortured), problems making a living (such as lose your job, can’t go to work, or can’t do farming), extortion, forceful destruction of property (such as burning your house or other possessions), and other. Responses were coded into two dichotomous variables of interest, 1) physical danger was the biggest worry or 2) problems making a living were the biggest worry. These two variables do not measure amounts or frequency of fear (i.e. how much people worry), but are comparisons of whether a respondent was more concerned about one particular problem such as physical danger over another problem such as livelihoods (i.e. what people worry about). The other possible responses, worry about extortion or forceful destruction of property, were not analyzed due to the low frequency of respondents who identified these as their biggest worry (4% in each case). About 25% of respondents were not able to choose a single issue that was their biggest concern⁷. We use these two variables (physical danger is biggest worry and livelihood is biggest worry) as outcomes in two different regression models. For each of these, all other answer choices were coded as “0”; for example for physical danger is biggest worry, the answer choices for livelihoods, extortion, property damage, and cannot choose biggest concern are all coded as “0”.

Next, respondents were asked, “During the conflict, in general how worried were you about physical danger (such as getting killed or injured) to you or your family members? Would you say that you worried about it every day, once or twice a week, once or twice a month, or less than once a month?” The use of emotional terms, such as ‘worry’ and ‘concern’ in these questions appropriately address the emotion of fear that we are seeking to address (as opposed to a cognitive judgment of risk) and are similar to successful fear of crime survey questions used in the United States (Ferraro and LaGrange 1987; Lalli and Savitz 1976; Warr 1984; Warr and Stafford 1983). Furthermore, although fear of violence may be a key mechanism linking exposure to armed conflict with adverse mental health outcomes, the emotional response of fear that we measured in this study is not a proxy for mental health disorder. Tests of correlations between fear of violence and PTSD and depression from the same study were not large or statistically significant⁸.

Note also that this question addressed the respondent’s fear for their family members as well as themselves. Answers to this question were coded into an ordinal variable, where “1” indicated the lowest frequency of fear (less than once a month) and “4” indicated the highest frequency of fear (every day). This is a measure of the overall fear of danger that respondents experienced, or how much people worried about danger.

As shown in Table 2, the mean score to the frequency of fear question was 2.13, somewhere in between once or twice a month and once or twice a week. Women had a higher frequency of fear, with a mean of 2.30, compared to 1.90 for men. This sex difference is statistically significant with p<0.01. With reference to the biggest worry questions, 46% of people stated that danger of physical harm or injury was their biggest worry, again with a higher percentage among women than men. Alternately 21% stated that livelihood problems were their biggest worry, with a higher percentage of men falling in this category.

Table 2.

Descriptive statistics for fear of violence and traumatic experiences

		Full sample			Men		Women
	Range	Mean	Std dev	Correlat'n with freq of fear	Mean	Std dev	Mean	Std dev
Fear and worries during conflict
Frequency of fear	(1–4)	2.13	1.17	1.00	1.90**	1.15	2.30**	1.15
Physical danger is biggest worry	(0,1)	0.46	0.50	0.18	0.40*	0.50	0.51*	0.50
Livelihood is biggest worry	(0,1)	0.21	0.41	0.31	0.24	0.43	0.19	0.39
Property damage is biggest worry	(0,1)	0.04	0.20	0.28	0.05	0.22	0.04	0.19
Extortion is biggest worry	(0,1)	0.04	0.19	− 0.10	0.01	0.11	0.06	0.23
Traumatic experiences
Conflict related trauma – summary	(0–9)	0.70	1.10	0.13**	1.02***	1.41	0.48***	0.75
Conflict related trauma – ever	(0,1)	0.67	0.47	0.14*	0.75***	0.44	0.62***	0.49

Statistically significant differences between men and women indicated:

^*p<.05

^**p<.01

^***p<.001 (one-tailed tests)

There is notable association between these fear measures. For example, as shown in Table 2, the polychoric correlation between frequency of fear and physical danger as biggest worry is 0.29, for livelihoods as biggest danger is 0.18. The polychoric correlation between frequency of fear and a categorical measure of biggest fear (with danger, livelihoods, etc. as different categories) is 0.41. This association is not unexpected, but the moderate (and not extreme) nature of the correlations indicate that these variables are measuring different (if related) concepts.

4.2 Traumatic experiences

Respondents were questioned about 16 different traumatic experiences they had ever experienced⁹. In this study, we use a summary measure of the number of different types of the following conflict-related traumatic events that a respondent had experienced¹⁰. As shown in Table 2, the summary measure ranged from experience of zero to nine different kinds of traumatic events. 67% of the study population had experienced at least one of these events and the average number different kinds of events experienced (the summary measure) was 0.70. A larger percentage of men had experienced any traumatic events, 75% compared to 62% of women. Men also experienced more than twice as many different kinds of events than women, with 1.02 on the summary measure, compared to a much lower 0.48 for women. The sex difference on both of these variables is statistically significant with p<0.001. As shown in Table 2, we can also see that traumatic experience is positively correlated with fear of violence, but with coefficients of 0.13 and 0.14, this correlation is relatively weak, indicating that other factors must play a key role in influencing fear.

4.3 Demographic and Family Characteristics

Respondents were also asked a variety of questions about themselves and their families. We use a dichotomous measure of sex which is coded ‘1’ for women, and ‘0’ for men. To address our hypotheses related to age-specific social responsibility and risk of experiencing violence, we create three categories for age. These include 17–24 years old, 25–45 years old, and older than 45 years. Two categories are used for marital status: never married and currently married¹¹. Finally, we use a dichotomous measure for large households, which includes those who live in households with six or more people.

In addition to the key independent variables that relate to social roles and responsibilities, we include two other controls in all our models. Caste/ethnicity is important because it affects almost everything in social life in Nepal. Although we recognize that Nepal’s ethnic groups are extraordinarily diverse, for the purposes of our analyses, we divided them into four comparative groups – Brahmin\Chhetri and Newar, Dalit, Hill Janjati and Terai Janjati (for detail see Bennett et al. 2008). We also control for education, or basic literacy and numeracy, with a dichotomous measure. Respondents who had completed at least three years of school were assumed to have these basic skills, whereas those with less than three years were assumed not to have these skills.

4.4 Analytic Methods

We use a series of regression models to predict the three outcomes- frequency of fear, danger was the biggest worry, and livelihood problems were the biggest worry¹². We first use ordinal logistic regression to predict frequency of fear, which is an ordinal variable with four possible outcomes on the full sample. Then we predict frequency of fear on men and women separately to test whether specific demographic and social factors have a greater or lesser effect on fear for each sex. These results are presented in Table 3. Next, we use logistic regression to predict the two dichotomous outcomes, physical danger was the biggest worry and livelihood problems and present the results in Table 4. We tested these two outcomes on male and female samples separately, but found no substantial differences by sex.

Table 3.

Ordinal regression models, predicting frequency of fear.

	Model 1 Full sample	Model 2 Men	Model 3 Women
Traumatic Experiences
Conflict-related trauma (# of diff events experienced up to 4)	1.45*** (3.58)	1.47** (2.88)	1.47* (2.19)
Demographic Characteristics
Sex (female)	2.14*** (3.42)
Age
Young (age 17–24)	Reference	Reference	Reference
Middle (age 25–45)	0.49* (2.21)	0.16* (2.08)	0.65 (1.21)
Older (age 45+)	0.41* (2.26)	0.14* (2.12)	0.53 (1.26)
Family Characteristics
Large household	1.24 (1.10)	1.17 (0.50)	1.29 (1.00)
Marital Status
Never married	Reference	Reference	Reference
Married	2.20* (2.04)	5.15* (1.76)	2.17^ (1.63)
Ethnicity
Brahmin/Chhetri or Newar	Reference	Reference	Reference
Hill Janjati	0.98 (0.07)	1.64 (1.25)	0.70 (1.02)
Dalit	1.07 (0.21)	1.57 (0.90)	0.92 (0.22)
Terai Janjati	1.14 (0.50)	1.53 (0.98)	0.94 (0.17)
Education
Education (3+ years completed)	1.34 (1.23)	1.03 (0.08)	1.77* (1.81)
−2 log likelihood	966.429	367.632	586.659
No. of Observations (people):	384	161	223

Odds ratios presented. Z-statistics in parentheses.

^{^}p<.10

^*p<.05

^**p<.01

^***p<.001 (one-tailed tests)

Table 4.

Logistic regression models, predicting comparative concern for physical danger and livelihood problems.

	Model 6 Physical danger is biggest worry	Model 7 Livelihood is biggest worry
Traumatic Experiences
Conflict-related trauma (# of diff events experienced up to 4)	1.13 (1.02)	1.10 (0.68)
Demographic Characteristics
Sex (female)	1.74* (2.26)	0.64^ (1.52)
Young (age 17–24)	Reference	Reference
Middle (age 25–45)	0.69 (1.01)	1.39 (0.72)
Older (age 45+)	0.60 (1.14)	0.63 (0.80)
Family Characteristics
Large household	1.15 (0.65)	0.71^ (1.30)
Marital Status
Never married	Reference	Reference
Married	1.10 (0.22)	0.48^ (1.38)
Ethnicity
Brahmin/Chhetri or Newar	Reference	Reference
Hill Janjati	1.43 (1.25)	0.49* (1.73)
Dalit	0.62^ (1.41)	2.98*** (3.07)
Terai Janjati	0.50* (2.27)	1.20 (0.54)
Education
Education (3+ years completed)	1.73* (2.06)	0.84 (0.53)
−2 log likelihood	499.50	372.389
No. of Observations (people):	384	384

Odds ratios presented. Z-statistics in parentheses.

^{^}p<.10

^*p<.05

^**p<.01

^***p<.001 (one-tailed tests)

For the outcome “physical danger is biggest worry” all other answer choices are included in the reference category, including livelihood, property damage, extortion, and cannot choose biggest worry. For the outcome “livelihood is biggest worry” all other outcomes are also included in the reference category.

5. RESULTS AND DISCUSSION

5.1 Frequency of Fear

Experience of traumatic events, sex, age, and marital status all had statistically significant effects on the frequency with which individuals worried about physical danger during the conflict. As shown in Model 1, the experience of conflict-related traumatic events increased the frequency with which individuals worried about physical danger; for every different kind of event an individual had experienced, the odds of being in a higher fear category were 1.45 higher. This supports our hypothesis 1 (experience of trauma will increase fear of violence). Looking at Models 2 and 3, we find coefficients of similar magnitude for men and women, suggesting that experience of conflict-related traumatic events influenced men’s and women’s fear similarly.

Sex also influenced frequency of fear as expected. With a coefficient of 2.14 in Model 1, women were more than twice as likely to be in a higher fear category than their male counterparts, independent of traumatic experience and other factors. Although men experienced more traumas (as shown in Table 2), this sex effect on frequency of fear is independent of traumatic experience which is controlled in Model 1. This is consistent with our hypothesis 2 (women will experience more fear of violence than men). It is particularly interesting, because recall that during this conflict, men were at greater risk of conscription into military forces and faced greater or at least equal risk of randomly experiencing violence as women. Yet we still find that women generally experienced more fear than men.

Model 1 also suggests that the influence of age on fear of violence in this study is substantial. With odds ratios below 1.0 (0.49 for the 25–45 year old and 0.41 for the 45+ age group), our results indicate that the youngest age group (17–24) has the highest fear. Models 2 and 3 show that this age is stronger for men (as would be expected), but still substantial for women.

This result supports our hypothesis three, that the effect of age on fear is driven primarily by the increased exposure to violence and targeting of younger people. In particular, it was most often young men who were targeted for violence and conscription into arms. As such, we would expect the age effect on fear to be largely driven by men, as it is here. It also suggests that the reason that women experience higher levels of fear than men is not because they are at greater risk of experiencing violence. If that were the case, we would expect to see a significant age effect for women. Instead, these results indicate that male fear (which is greatest at youngest ages) might at least partially come from men being at greater risk of violence and conscription. Women’s fear on the other hand, which we find is at similar levels for all female age groups, might be a product more of altruistic fear for others in their households, perceived vulnerability, and perceived or real lack of agency.

Being married had positive effects on fear. Living in a large household also had positive effects, but the lack of statistical significance for the large household variable gives us little confidence in that result. Those who were married were more than twice as likely to be in a higher fear category compared to those who were unmarried. These results support hypothesis 4 (married individuals will report more fear), but not hypothesis 5 (those living in larger households will report more fear). While these results are opposite of what is expected in the fear of crime literature, we argue that the results need to be contextualized. The meaning of marriage differs between Nepal (this study) and the U.S. and Europe (where most studies of fear of crime are based). In the fear of crime literature, marriage is often equated with increased social support, and social support serves as a protector against the development of fear. Alternately in Nepal, marriage appears to increase both social responsibilities and social support, but we believe that the increase in responsibility is more significant than the increase in support. As such, it is the social responsibility mechanism that connects marriage to increased fear of violence in Nepal, and it is the social support mechanism that connects marriage to decreased fear of crime in the U.S. and Europe.

The sex differences in results for marriage are notable. Marriage had a much stronger effect on men (Model 2), with a statistically significant coefficient of 5.15. This means that men who were married were more than five times more likely to be in a higher fear category than their unmarried male counterparts. Women, on the other hand, experienced a smaller (odds ratio = 2.17) effect of marriage on fear, which was only marginally statistically significant. This difference in the effect of marriage for men and women suggests gendered patterns of altruistic fear similar to what we find in the fear of crime literature. Particularly, the increased fear of married men might indicate spousal fear for their wives.

These results generally support our hypotheses and together suggest that social roles and responsibilities have important influences on the development of fear of violence during armed conflict. First, in this context, women, younger adults, and married adults all have more responsibilities for others’ welfare in addition to their own. Our analyses show that all of these groups experienced significantly more fear than their less encumbered counterparts, an indication of altruistic fear.

Second, the sex results in this study are similar to those for fear of crime in high income countries. As suggested in the fear of crime research, this gender difference in fear might also be a result of women’s perceptions of their vulnerability (LaGrange and Ferraro 1989; Franklin and Franklin 2009; Smith and Torstensson 1997; Stanko 1997; Warr 1984; Snedker 2012). This is not to say that women are more vulnerable, but in highly gender stratified patriarchal societies such as Nepal, they are often socialized to believe that they are more vulnerable. The case of Nepal is particularly instructive here for understanding the fear of crime literature. In that literature, based primarily in the U.S., women’s greater sense of vulnerability is linked to the “shadow” of sexual assault whereby fear of rape as a "master offense" increases women’s fear of crime (Ferraro, 1995, 1996; Fisher and Sloan, 2003; Lane and Meeker, 2003; Warr 1984). While women are more likely than men to be victims of sexual assault, they are less likely to be victimized by other crimes, and also appear to experience greater vulnerability. Thus it is difficulty to fully unpack the extent of women’s greater fear of crime due to perceived vulnerability or to objective greater risk (of sexual assault). During the conflict in Nepal however, men were at greater risk of violence, but women experienced more fear. This allows us to suggest that Nepali women’s fear must be more affected by perceived vulnerability and/or altruistic fear for family members.

5.2 Biggest Worry

Turning to comparisons of types of fear, we now address whether respondents reported that their biggest worry was physical danger or livelihood problems. As shown in Table 2, 40% of men and 51% of women cited physical danger as their biggest concern. Alternately, lower than physical danger but still a substantial proportion of respondents reported that livelihoods were their biggest concern, 24% of men and 19% of women.

Models 4 and 5 in Table 4 present the results of models testing the demographic and social factors that influenced which problem respondents were most concerned about. Notably, although experience of traumatic events increased the frequency of fear (as shown in Model 1), it had no significant effect on the kind of problems respondents were most concerned about. Traumatic experience was neither statistically significant in Model 4 as a predictor of comparative concern for physical danger, nor in Model 5, as a predictor of comparative concern for livelihood problems. This result adds nuance to our argument. While trauma increases fear, it does not influence fear for one type of worry over another, suggesting different dimensions in the development of fear. There are individual determinants of fear levels and types, but those determinants may differ.

Sex had a large effect on the type of concerns individuals had with regard to the conflict. As shown in Model 4, women were about one and a half times more likely than men to state that danger was their biggest concern. On the other hand, as shown in Model 5, women were about 36% less likely than men to report that livelihood problems were their biggest concern. These results support hypothesis 2 and provide further evidence that gender roles and responsibilities are key factors in the development of fear and worry. In this case, women, who are responsible for the well-being of others in the family were more likely to worry about danger to themselves and their family, again, evidence of altruistic fear of danger. On the other hand, men, who are generally responsible for livelihoods in Nepal, were more likely to worry about livelihood problems.

Alternately, age had little statistically discernable effect on the primary concerns for livelihood or danger during the conflict. We find negative odds ratios for older groups in Model 4, but these are marginally and not statistically significant, giving us little confidence in our results here and suggesting high variance in the relationship between age and comparative fear of danger. Younger adults were targeted for violence and conscription to a greater extent. However, they are also generally responsible for the financial well-being of the family. This is likely the reason we find little effect of age on primary comparative concerns, which is instead largely determined by gender and family factors.

Marriage had a negative and marginally significant influence on fears for livelihoods. Respondents who were married were about half as likely to report that livelihood problems were their biggest concern. We find similar negative and marginally significant effects of household size on livelihood worries. This difference in the effects of marriage and household size on the biggest worry is as expected; we described being married and living in a large household as similarly increasing the social responsibility of the individual. As such we would expect them to have similar effects.

5.3 Qualitative Analysis

As we discussed above, the data for our qualitative analysis comes from in-depth interviews with residents of Chitwan. We use the insights from the in-depth interviews to contextualize our quantitative findings, particularly about gender roles.

Our in-depth interviews show that both men and women expressed fear for economic livelihood but it was talked about in gendered ways. For example, one male resident stated that “I was more worried about my family. I couldn’t imagine how they managed everything in my absence while I, the breadwinner, was imprisoned” (emphasis added). This fear was echoed directly by another man, “I had to look after the children and feed them.” Women also expressed concerns about daily economic issues, but it was in terms of their husband’s ability or inability to provide: “I was afraid if we would go without food. I didn’t have any job. The whole family depended on the income of my husband only. I was worried how we could manage to buy everything while living in the urban area.” Several women lost their husbands in the wake of the armed conflict. The economic impact was painfully expressed in the words of one woman, “I was dependent on my husband. But now I have got the great burden of responsibility” and similarly by another women, “But, after his death, nobody is there to take care of us. Nobody is helping us.” In contrast women were “much more worried about the children” and their spouses, “I was most frightened about my husband. I wondered whether he would get released [from prison] or not. I was afraid that he would be killed...”

Respondents also expressed how men played the role of protector, which impacted women’s great personal fear. “Whenever there was no one at home, I used to get afraid. But whenever my husband was there at home, I didn’t feel afraid.” In interviews, both men and women expressed fear of violence for oneself. For some women, like in the case of US fear of crime research, this fear was directly tied to fear of sexual assault. One woman stated “I feared if they would rape me. I was terrorized all the time. There was not a single day when I got a peaceful sleep.” Respondents expressed that these fears were more pronounced at night and when alone, “I was staying alone there… I spent the night at risk.”, again suggesting the protector role of men. While our interviews suggest that men were generally considered to inhabit the protector role, women tended to express greater altruistic fear for both children and spouses. This contrasts with U.S.-based research that finds women’s altruistic fear is directed at their children (parental fear) and men expressed fear for their partners/wives (spousal fear).

6. CONCLUSION

A comprehensive understanding of the consequences of armed conflict on civilian populations is necessary to appreciate their experiences and well-being during conflict, but also to recognize and identify the long-term impacts of armed conflict in post-conflict societies, including enduring changes in mental health. Towards this end, this article provides key insights into how individuals process their experiences living through armed conflict, resulting in significant and chronic fear of violence and concerns for their livelihoods. This is one of the first articles of which we are aware that includes a theoretical discussion and quantitative empirical examination of the determinants of who is more affected by fear during armed conflict.

Results from this study show that chronic fear of violence was prevalent in Nepal during the armed conflict. However, although fear of violence was significant, violence and physical danger were not the most important concern for a sizeable proportion of people during the conflict. Armed conflict can disrupt the livelihoods that support individuals’ and families’ daily living and this disruption was the greatest concern for many people in our study. This aspect of concern is often disregarded by researchers and policy officials alike.

We also show that all people do not experience fear of violence equally during armed conflict. Those who had greater responsibilities for others’ welfare, were more at risk, or might have perceived themselves to be more vulnerable, generally experienced more fear and worry. In other words, fear of violence disproportionately affected those who were already more vulnerable and those who shouldered more responsibility for others, possibly further increasing existing social inequalities. As with many quantitative studies, we use measures of sex, marital status, age, and household size to represent gender roles and social responsibility. While these are likely quite good proxies for the concepts of roles and responsibilities in the Nepali context, additional qualitative research exploring these connections would likely yield interesting insights and nuance to this area of study.

Results in this study are comparable in many ways to research on fear of crime in high income settings during periods of relative political peace. For example, in both cases, sex was amongst the strongest predictors of fear, with women experiencing significantly more fear than men (Franklin and Franklin 2009; Rountree and Land 1996b; Warr 1984). However, the results differed from early research on fear of crime and age. In Nepal during the armed conflict, younger adults unambiguously experienced more fear; alternately in many studies older people in the United States and Europe experience more fear of crime. More recent studies, especially those conducted outside of the U.S., suggest that fear of crime does not increase with age (Chadee and Ditton 2003) and that younger individuals have higher levels of fear (Liu et al. 2009), consistent with the findings presented here.

Our findings highlight the importance of context as well as measurement of fear. In terms of measurement, this study questioned respondents on their fear for their own as well as other family members’ safety. Our findings are consistent with altruistic fear – fear for others – that is identified in the fear of crime research. In terms of context differences, studies in the United States argue that amongst the reasons that older people are more fearful is because they are socialized to believe that they are more vulnerable (Warr 1984). In the context of Nepal, older people and those who were married were less preferred as recruits by both warring parties. Further, older age is largely correlated with less exposure to violence, less time spent outside the home, and fewer family responsibilities. As such, older people in Nepal may feel less at risk, less vulnerable, and less responsible for others’ welfare, which then translates into less fear of violence.

In addition to drawing heavily on the fear of crime research, this article also contributes to that area of study. We find that a significant proportion of our respondents reported that they were more concerned about livelihood and other financial consequences of armed conflict than about danger. We also find systematic variance in who is more likely to report danger or livelihoods as their biggest concern. This topic, of comparative concern for things besides danger, is largely absent from the fear of crime literature. Yet, it is likely that fear of livelihood and financial problems exists in U.S. neighborhoods with high crime, just as in Nepal during the armed conflict.

One of the main differences of our study to the broader fear of crime research is the effect of marriage on fear of violence. We believe that the reason our marriage results are almost exactly opposite to those in studies of fear of crime is because of context. In Nepal, it is likely that marriage increases social responsibilities to a greater extent than it increases social support. Thus it is not necessarily that different mechanisms are influential, but that the context determines which mechanisms are more and less relevant to the relationship under study. Given the large role that context might play, we recommend that our understanding of fear of crime would benefit from careful attention to cultural context in studies based in areas of the world outside the U.S. and Europe.

Another way in which this study contributes to the academic literature is through a detailed exploration of fear of violence. Fear of violence might be one way in which armed conflict influences other outcomes that are of serious academic and policy concern. These include physical and mental health, treatment seeking behaviors, marriage, childbearing, and migration. Moreover, differences in levels and types of fear might help to explain differences in the effects of armed conflict on these other outcomes. Our study questioned respondents about fear of violence three years after the conflict ended, providing evidence that fear of violence can persist in the long-term. While a permanent cease in macro-level violence will undoubtedly ease individuals’ fears to some extent, it is also likely that some of these fears for the dangers of the macro-level context will remain for many years. This could have significant impacts on physical and mental health, and many other behaviors.

This study might also be relevant to policymakers. For example, programs to support civilians in post-conflict situations might be more effective if designed to consider multiple fears surrounding conflict (livelihood problems and property destruction as well as danger). This study also highlights the potential long-term nature of fear, which might contribute to interventions addressing mental health symptoms and disorders well into a post-conflict period.

There are limitations to this study and our results. Our sample size (at 384 people once the widowed and divorced individuals were deleted) is not large and precludes us from more fully investigating certain findings. Similarly, we are not able to directly measure or test the mechanisms we suggest. For example, we use marriage, household size, age, and sex to measure social roles and responsibility, but more direct measures and testing of these concepts could more clearly identify the mechanisms connecting marriage, age, and sex to differences in fear of violence. While our measurement of fear of violence and biggest worry shows strong and theoretically supportable results, we gathered this information three years after the conflict ended. Despite some limitations, this paper represents a strong beginning to understanding fear of violence during armed conflict, guiding future studies on this topic.

Future research in this area would benefit from addressing some of the limitations of this study. Data collection during a conflict, more detailed and direct measurement of theoretical mechanisms, and larger sample sizes would all create the ability to test the relationships in this study in more detail and nuance. Additional qualitative analysis of in-depth data will likely yield more insights to these questions as well. Finally, data collection at multiple points in time to create a longitudinal record of fear of violence would add an entirely new dimension for investigation of these questions. Future studies can build on the current study to increase our understanding of fear and the many and broad consequences of armed conflict on civilian populations. This study highlights similarities between fear of violence during armed conflict and fear of crime in general. Our findings yield additional and important insights to the existing literature on fear of crime.