The internal structure of crime as a problem to the private self

With this knowledge of the underlying elements and inter-relationships of the sub-concepts of the fear of crime, we will try to explain the paradoxical observation made earlier in respect of the descriptive statistics (§3.3): the respondents who experienced problematic crime in their neighbourhood most frequently and who had fallen victim of crime the most (Amsterdam) thought they had the smallest chance of falling victim of crime and vice versa (Zaltbommel). This observation raises the following question: how does crime –experienced as a nearby problem in one’s neighbourhood and/or distant in society – turn into a significant threat to the private self, or does it not? To answer this question, we will zoom into the internal structure of crime experienced as threat to the private self.

Observations Ρ χ²/df CFI TLI RMSEA CD

Total 2786 .00 14.70 .89 .86 .07 .91

Amsterdam 559 .00 3.66 .90 .87 .07 .71

Hilversum 1189 .00 7.19 .88 .85 .07 .77

Zaltbommel 1034 .00 5.86 .89 .86 .07 .93

Note. CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root mean square error of approximation; CD = coefficient of determination Tab. 28 – Goodness-of-fit statistics, internal structure of crime as a problem to the private self.

Fig. 23 –Internal structure of crime as a problem to the private self, standardised coefficients.

Crime - situated either in the psychologically nearby context of the neighbourhood or in the distant context of Dutch society – turns into a significant ‘problem to the private self’ (tab. 28 & fig. 23) through one’s ‘estimated chance of personal

victimisation’ (λ=.47). In particular, the assessment of nearby crime - in the form of the ‘estimated occurrence of crime in the neighbourhood’ (λ=.26) influences one’s

‘estimated chance of personal victimisation’. The related (Ψ=.19) assessment of distant crime - in the form of the ‘estimated occurrence of crime in society’ - is of less influence (λ=.12). One’s assessment of ‘vulnerability’ also has considerable influence over the estimation of one’s ‘chance of personal victimisation’ (λ=.08). This concept of ‘vulnerability’ is theoretically held to be a product of ‘gender’ and 'age‘, because generally women and the elderly assess themselves to be more vulnerable to fall victim of crime. We now observe significant covariance between

‘vulnerability’ and the ‘estimated occurrence of crime in society’ (Ψ=.09). This is in line with earlier descriptive observations (see section 7.2.2), that older respondents worried more about distant crime in specific and that women are more sensitive to crime than men in general. The cognition of ‘crime as a problem to the private self’

KEEPING TROUBLE AT A SAFE DISTANCE

leads respondents to report ‘avoidance behaviour’ (λ= .26), but also “activates”

‘psychological defense mechanisms’ to a stronger magnitude (λ= .43). We already know from the previous sections (sections 7.4.1 & 7.4.4) that these ‘psychological defenses’ strongly suppress the larger concept of ‘personal fear of crime’ (fig. 2 & 5). 7.4.6 Explaining municipality differences

Having observed the internal structure of the cognition of ‘crime as a problem to the private self’, we can now search for more detailed explanations for the observation made earlier: that the respondents who actually experienced nearby problematic crime the most (Amsterdam) paradoxically thought they had the smallest chance of falling victim of crime themselves and vice versa (Zaltbommel). The goal is to search for significant differences⁴⁹ for the respondents of Amsterdam compared to respondents from Hilversum and Zaltbommel, which could explain this

observation.

In this context, it is interesting to look at path-diagram differences for the ‘personal and situational fear of crime model’ (tab. 26) and the ‘the internal structure of crime as a problem to the private self’ model (tab. 29).

As shown in the table on the previous page (tab. 29), at face value the respondents from Amsterdam show the strongest suppressive function of ‘psychological defenses’ to ‘personal fear of crime’. Another initial impression is that they also

49 Here we used the Likelihood Ratio (LR) Test, calculating the difference between:

(I) the χ²-value of the SEM with constraints for the path coefficient of our interest in all municipalities and (II) the χ²-value of the SEM with constraints for the path coefficient of our interest only in the municipalities of Hilversum and Zaltbommel (See Bollen 1989:292; Ullman &

Bentler 2013:675-676 for details).

Amsterdam Hilversum Zaltbommel Personal fear of crime 

Situational fear of crime .08 .10* .02

Psychological defenses -.54* -.49* -.44*

Authoritarian sentiments .04 -.00 .11

Societal conservatism .26* .35* .26*

Situational fear of crime 

Dispositional fear .52* .42* .52*

Authoritarian sentiments .22* .11* -.04

Societal conservatism .08 .22* .22*

Avoidance behaviour 

Situational fear of crime .43* .67* .59*

Personal fear of crime .11* .06 -.00

Dispositional fear .31* .13* .16*

*Significant, p < .05

Tab. 29 – Path coefficients for the three municipalities for the personal and situational fear of crime model.

MEASURING ‘THE FEAR OF CRIME’

reported most ‘avoidance behaviour’ due to ‘personal fear of crime’. But Likelihood Ratio tests showed no significant difference for the functions of both ‘psychological defenses’⁵⁰ and ‘avoidance behaviour’⁵¹ for the Amsterdam respondents compared with respondents from Hilversum and Zaltbommel.

Amsterdam Hilversum Zaltbommel estimated chance of personal victimisation 

estimated occurrence of crime in neighbourhood .36* .26* .30*

estimated occurrence of crime in society .00 .12* .09*

Vulnerability .08 .08 .08*

crime as a problem to the private self 

estimated chance of personal victimisation .50* .50* .44*

Psychological defenses 

crime as a problem to the private self

.46* .40* .46*

Avoidance behaviour 

crime as a problem to the private self .34* .26* .20*

*Significant, p < .05

Tab. 30 – Path coefficients for the three municipalities for the ‘crime as a problem to the private self’ model.

When we look at the path coefficients for the ‘crime as a problem to the private self’ model, at face value it seems that Amsterdam respondents show the strongest activation of ‘psychological defenses’ and ‘avoidance behaviour’ as a result of the cognition of ‘crime as a problem to the private self’ (tab. 30). Likelihood Ratio tests for the significance of these differences showed that the respondents from Amsterdam did have a significantly stronger⁵² activation of ‘avoidance behaviour’

as a result of the cognition of ‘crime as a problem to the private self’. But the activation of ‘psychological defenses’ as a result of the same cognition was not significantly stronger⁵³.

This leads us to the conclusion that the respondents from Amsterdam reported a stronger activation of ‘avoidance behaviour’ due to the cognition of ‘crime as a problem to the private self’, compared with the respondents of Hilversum and Zaltbommel. The respondents for all three municipalities showed a similarly strong activation of ‘psychological defenses’ and a similar suppressive function of such defenses to ‘personal fear of crime’.

50 LR = 2.87 < χ²crit (1) = 3.84; p= .05.

51 LR = 3.00 < χ²crit (1) = 3.84; p= .05.

52 LR = 5.12 > χ²crit (1) = 3.84; p= .05.

53 LR= 2.06 < χ²crit (1) = 3.84; p= .05.

KEEPING TROUBLE AT A SAFE DISTANCE

These insights provide an integrated explanation for our paradoxical observation that respondents from Amsterdam assessed the smallest chance of falling victim of crime themselves, while: (I) Amsterdam has the highest crime rates by far; (II) respondents from Amsterdam actually fell victim of crime the most; (III) respondents from Amsterdam assessed of the occurrence of crime in the neighbourhood to be the most frequent; and (IV) respondents from Amsterdam experienced crime to be a neighbourhood problem the most.

The paradoxical observation can be explained as respondents from Amsterdam showed: (I) a basic suppressive function of psychological defenses to personal fear of crime; (III) reported most avoidance behaviour as a consequence of the

cognition of ‘crime as a problem to the private self’; and (IV) showed the strongest activation of avoidance behaviour as a result of the cognition of crime as a problem to the private self. In this way a psychologically nearby perceived threat of crime becomes cognitively neutralised.