412
Views
4
CrossRef citations to date
0
Altmetric
Articles

Correlates of Perceived Insecurity: Evidence from Pakistan

&
Pages 488-504 | Received 21 Aug 2016, Accepted 24 Jul 2017, Published online: 09 Aug 2017
 

Abstract

This study explores the correlates of perceived insecurity among the households in Pakistan. For this purpose, data from the third round of Pakistan Panel Household Survey are merged with terrorist incidents taken from Global Terrorism Database. The results illustrate that objective risk, signified by violence in the district of residence, and victimization play an important role in the formation of perceived insecurity. Moreover, males and residents of rural areas feel more insecure than females and urban residents, respectively. Furthermore, spatial variation in violence indicates that terrorist attacks in the first-order contiguous districts (i.e. the immediate neighbors) are also associated with subjective insecurity. These are crucial findings because behavioral changes, such as perceived insecurity, could force the households to make suboptimal investment decisions. More importantly, since changes in behavioral parameters may be highly persistent, this study cautions that violence in Pakistan may have potentially long-term impacts on social welfare.

JEL Codes:

Acknowledgments

We would like to thank Wayne Gray, Marc Rockmore, Dhanushka Thamarapani, Nadeem Iqbal, Beenish Javed, Naeem Khan, Yasmin Abdul Wahan and the three honorable referees for their valuable comments. The usual disclaimer applies.

Notes

1. Criminology literature has identified several correlates of perceived insecurity. These include gender, age, poverty, residence, neighborhood disorder, lack of social integration, frequent racial conflicts, poor street lighting, vandalism, dilapidated buildings and drunks loitering on street corners (see, for example, Pain Citation1995; Hollway and Jefferson Citation2000; Heidensohn Citation2002; Kanan and Pruitt Citation2002; Zedner Citation2002; Wiles, Simmons, and Pease Citation2003).

2. The word ‘agency’ is used in geographical sense and is equivalent to a district in the settled areas. There are seven agencies in the FATA in Pakistan; Khyber, Bajaur, Mohmand, Kurram, Orakzai, South Waziristan, and North Waziristan.

3. Pakistan Economic Survey (Citation2015).

4. It is important to note that the temporal variation in perception of insecurity is also important and interesting to investigate in the specific case of Pakistan. Sadly, in the first two waves of PPHS, data is not collected on perceived insecurity.

6. It is important to mention that even if the attacks are not indiscriminate, the probability of victimization is random. We check this using a Logit model by regressing probability of victimization on individual and household characteristics and found all variables to be insignificant. These results are available upon request.

7. It is, however, important to mention here that PPHS data is not weighted. Hence, the generalization of results should be carried with caution.

8. The discussion on the rational for the selection of these districts as a representative sample of the entire country is provided in Nayab and Arif (Citation2014).

9. Since the entire analysis is based on data from head of the household, therefore, we do not include a control dummy for this status. It will dropout from the regression due to lack of variation.

10. Using violence in the model also serve the purpose of separating the effect of pure terrorist activities from non-terrorist criminal activities (also done by terrorist groups) on the perception of insecurity.

11. The question asked was about being a victim (both physical injury and property damage) in the last 12 months.

12. This is not to say that women and kids have never been killed in random terrorist attacks. This point is about specifically targeting women and kids. Terrorists did not explicitly target kids and women as it is against the cultural norms and hence would result in loss of support from their own sympathizers. The attack on Malala Yousafzai and the school attack are clear examples of that.

13. However, these attacks may be less deadly than the ones in urban regions.

14. However, these differences in magnitudes are not statistically significant.

15. We are thankful to the anonymous referee for this suggestion.

16. That is, all those observations which are in the first category are coded as 1 while observations in rest of the categories are coded as zero.

17. This will require using autofit command along with gologit2 in Stata.

Log in via your institution

Log in to Taylor & Francis Online

PDF download + Online access

  • 48 hours access to article PDF & online version
  • Article PDF can be downloaded
  • Article PDF can be printed
USD 53.00 Add to cart

Issue Purchase

  • 30 days online access to complete issue
  • Article PDFs can be downloaded
  • Article PDFs can be printed
USD 417.00 Add to cart

* Local tax will be added as applicable

Related Research

People also read lists articles that other readers of this article have read.

Recommended articles lists articles that we recommend and is powered by our AI driven recommendation engine.

Cited by lists all citing articles based on Crossref citations.
Articles with the Crossref icon will open in a new tab.