We look at a two-stage model where, in the first stage, every individual decides whether to become criminal and, if she or he chooses to take action, they decides just how much crime to exert in the next stage. We show the way the distance to the criminal leader affects both decision to become criminal (extensive margin) and the amount of crimes thereafter committed (intensive margin).
Data and empirical framework
We test our theoretical predictions using data from the National Longitudinal Study of Adolescent to Adult Health (Add Health) in america, which contains information on all students attending a random sample folks high schools in 1995. This dataset provides unique information on friendship networks by asking students to nominate up to ten friends from a school roster. In addition, it contains detailed information on juvenile delinquency, including 12 types of crime. To recognize criminal leaders in a manner that is exogenous to the network formation process, we define a criminal leader as an adolescent who has a degree of crime a lot more than three standard deviations above the median in the institution. The distance to the first choice is then calculated utilizing the (shortest) distance between any delinquent and the first choice in the social networking to which they belongs. Our identification strategy is founded on the actual fact that students choose their friends, and perhaps the friends of their friends, however, not beyond. The question we study in the empirical analysis is how being (randomly) located at a particular distance to the first choice affects a person’s criminal activities.
Figure 1 reports the correlation between your distance to a leader and crime activities in the raw data. The figure shows the common crime level for folks at different path lengths from the first choice, highlighting the unambiguous negative correlationbetween crime and the length to a leader.
Figure 1 Distance to the first choice and criminal activity
Results from our analysis indicate thatthe longer the (social) distance to the first choice, the low the criminal activity of a delinquent and the not as likely they’re to become criminal. Interestingly, even leaders far away of seven nodes from a delinquent still affect the latter’s criminal activities, although the result is weaker than that for someone located nearer to the leader. Regarding magnitude, for every one node upsurge in distance (i.e. one link from the leader), the full total number of crimes committed by each student decreases, normally, by 0.2 units (about 10% of a typical deviation). We then simulate the consequences of a policy looking to remove all criminal leaders from a school. We show that policy can, normally, reduce crime by 20% and reduce the individual probability of learning to be a criminal by about 10%.
Our results claim that criminal leaders are essential determinants of criminal activities. Quite simply, being socially near leaders matters. Consequently, targeting criminal leaders that are socially linked to other criminals could possibly be a competent policy for reducing crime.
Other researchers have suggested that targeting the so-called key playersin internet sites could possibly be another effective policy for reducing crime (Ballester et al. 2006, 2010, Zenou 2016, Liu et al. 2018). Key players are delinquents who, once taken off the network, generate the best decrease in crime. They are characterised by their centrality (or position) in the network (referred to as ‘intercentrality’).
The main element player policy, however, is founded on peer effects and the main element player is the person that reduces these peer effects the most. In this study, we voluntarily abstract from general peer effects and focus instead on an extreme peer, namely, the criminal leader. Unlike the main element player policy, where in fact the person targeted may be the person that gets the highest intercentrality, our study shows that key players could possibly be thought as the most active criminals who’ve a significant social influence in a network (i.e. the ‘key leader’ policy). Quite simply, according to your results, the most active delinquents who’ve high closeness centrality ought to be targeted.
Inside our dataset, key leadersare not necessarily key players, which means that both policies are distinct and capture different facets of criminal activities. Indeed, removing key (criminal) leaders leads to the average decrease in criminal activities around 20%, or a 33% fall for all those one link away. This sharp decrease in crime mostly occurs for all those significantly less than four links from the key leader. Consequently, in a dense network where in fact the key leaders are in most four links from the rest of the agents, removing key leaders works well at reducing crime.
Ballester, C, A Calvó-Armengol and Y Zenou (2006), “Who’s who in networks. Wanted: The main element player,” Econometrica 74: 1403-1417.
Ballester, C, A Calvó-Armengol and Y Zenou (2010), “Delinquent networks,” Journal of the European Economic Association 8: 34-61.
Egley, Jr, A and J C Howell (2011), “Highlights of this year’s 2009 National Youth Gang Survey,” OJJDP Juvenile Justice Fact Sheet, U.S. Department of Justice.
Levitt, S D and S A Venkatesh (2000), “An economic analysis of a drug-selling gang’s finances,” The Quarterly Journal of Economics 115: 755-789.
Liu, X, E Patacchini, Y Zenou and L-F Lee (2018), “Who’s the main element player? A network analysis of juvenile delinquency,” Unpublished manuscript, Monash University.
Reiss, A (1988), “Co-offending and criminal careers,” in N Morris and M Tonry (eds), Crime and Justice, University of Chicago Press, pp. 55-68.
Reiss, A and D P Farrington (1991), “Advancing understanding of co-offending: Results from a prospective longitudinal survey of London males,” The Journal of Criminal Law and Criminology 82: 360-395.
Warr, M (1996), “Organization and instigation in delinquent groups,” Criminology 34: 1l-37.
Zenou, Y (2016), “Key players,”, in Y Bramoullé, B W Rogers and A Galeotti (eds), Oxford Handbook on the Economics of Networks, Oxford University Press, pp. 244-274.