The interpretation of your results is overstated. It’s not clear that new dealers are killing established dealers. It could be the opposite pattern: Established dealers are killing new dealers, who, in turn, are replaced by new dealers. The statistical result would be what you found in this paper. This pattern makes as much sense as the idea you pose, though I appreciate your theoretical rationale for why it is what you claim. The way to discern which pattern actually applies requires knowing whether the homicide victims and perpetrators have a drug arrest record. If the perpetrators don’t have much or any drug arrest history, but the victims do, then your findings are supported. Otherwise, it remains an open question.
Page 11: To excuse something not included in your analysis, you wrote that “We can elicit no convincing theoretical rationale for why more or less drug enforcement would impact first-time and repeat offending differently.” Among other possibilities, new offenders have less expertise than seasoned offenders. Thus, seasoned offenders know how to better protect themselves, at least in theory. You should see the work of Claire Nee and Volkan Topalli, among others, on offender expertise. Given this, I’d like to know what happens to the results when the analysis includes “more or less drug enforcement.”
More important than that work to your paper are these articles. They have to be included in the review and their implications for your paper have to be spelled out: “Caulkins, Jonathan, Peter Reuter, and Lowell Taylor. 2006. Can supply restrictions lower price: Illegal drugs, violence and positional advantage. Contributions to Economic Analysis and Policy 5(Article 3).” And, “Levitt, Steven D., and Sudhir Alladi Venkatesh. 2000. An Economic Analysis of a Drug-Selling Gang’s Finances. The Quarterly Journal of Economics 115:755-789.”
Page 2: I found it sub-optimal to start a paper about systemic violence with a long paragraph about psychopharmacological and economic compulsive violence.
Your definition of non-drug homicides isn’t adequate. It should simply be defined as homicides that the police don’t think are related to drug trade. Your definition states that “non-drug homicides include homicides resulting from quarrels, robberies, assaults, etc.” But drug trade leads to quarrels, robberies, assaults, etc., so that can’t be used to distinguish drug homicides from non-drug homicides.