Update #2: Studying the Effects of Narcotics Control Policy on Enforcement Budgets


This is the second blog post update that I will be posting for my project.

Today marks the half-way point of my summer project investigating the narcotics policies of most nations on the planet and their relationships to enforcement spending within those same nations. The last 2 weeks have been filled with quite a bit of reading focused on scholarly articles and reports about both the variety and extent of different narcotics control regimes around the world. As expected, much of the world is largely homogenous when it comes to controlling the use of narcotics, though there are some notable exceptions in several countries, like Uruguay and Mexico, where governments have taken steps toward alternative models of control that move away from outright prohibition. The ideal scenario for this project involved becoming acquainted with the most recent changes in narcotics policy around the world, understanding how prohibition came to be the status quo in most nations, and then to collect and analyze data about how much these nations spend to enforce this status quo. Unfortunately, as it turns out, many nations, including industrialized, developed western countries, often do not publish or collect relevant data on enforcement spending or even on drug-policy related outcomes. My failure to find significant budget data on enforcement by nation is largely due to this fact, and it has led me to become a proponent of transparent budget standards for all nations and for this data to regularly be made available online for free. I suspect that many quantification problems in this field stem from this wide-spread data-neglect, which is quite horrifying considering the staggering costs in both dollars and human lives that are continuing to accrue from this set of policies. If any type of meaningful analysis of narcotics policies, status quo or otherwise, is going to be conducted, then international organizations and governments need to begin collecting and organizing budget information that can accurately account for the dispersal of funds toward specific policy objectives and not simply under broad categories like “health” and “defense” that may overlap on things like narcotics control, which in itself is an extremely broad policy category because of the sheer number of different substances that fall under the purview of it.

Despite these unfortunate circumstances, I have been persevering in the gathering of data. The Drug Freedom Index has proved to be very invaluable; I incorporated their 10 point scale that graded nations based on their drug laws (1 being the harshest and 10 being full legalization of all drugs) into my dataset. I also created my own variable from their data using instead a 4-category scale on which to score nations. These will now serve as the primary independent variables in my modeling. Of course, I needed to add control variables to account for other factors that might influence budgets. For this data I turned to databases created by the UN, World Bank, US Department of State, and the National Association of State Budget Officers. As of today, I have acquired an additional 11 variables within my dataset, bringing the total number of variables to 13. These include: The policy variables (10 and 4 point scales), unemployment rates, GDP per capita, health, education, and military spending as a percentage of GDP, Percentage surveyed who trust their government, the percent of US state budgets devoted to corrections, total population, and the amount of US money granted to other nations to combat the illicit narcotics trade. This last variable may prove to be the saving grace of this project as it helps quantify in a very similar way (at least from the United States’ perspective) how much these policies are costing to enforce. As of now and within the regressions that I’ve already run, it has been shown to be a very useful proxy variable that likely approximates the type of ideal variable that I had in mind when coming up with this research project. It should also be noted that most of the gathered data comes from the years 2012-2013, with only a couple of exceptions. My objective was to get the most recent data possible because of a number of recent shifts that have occurred within this policy field in just the last 24 months. Luckily I was able to find a good bit of recent data and there isn’t anything earlier than 2011.

It took a decent amount of time and effort just to organize the data within a single set in excel, but it seems to be paying off. The first models I investigated were OLS (ordinary least squares) models that used either the 10-point or 4-point policy variables as their primary independent variable. Within the model and holding the other variables constant, both policy variables proved highly significant at the p<.01 level and positively correlated with enforcement expenditure. Under my created variable (the 4-point scale), more variables were significant and it had an overall higher R-squared value than the 10-point scale. As a result, and following a reverse regression (policy as the dependent and enforcement as the independent), the 4-point variable appears to be more useful and will thus be used as the primary policy variable henceforth. I also made sure to use robust standard errors to eliminate any heteroskedasticity that might’ve been present

Overall, these initial results seem promising. A full statistical analysis of these and future models will be made in the final paper that will be written and compiled for this project, but for now it is sufficient to mention that the linear models have so far shown a number of interesting relationships between the variables and, despite the lack of good international budget data, the goals of this project are still very much achievable. However, reversing the variables on a hunch (as mentioned above) got me thinking. Maybe my hypothesized relationship between policy and enforcement spending runs the opposite direction. Maybe the capital being sent to other nations to fight the “War on Drugs” is what is driving policies to remain at the status quo and preventing alternatives from being instituted or even explored. This becomes even more plausible when you consider how developing nations may perceive these funds: taking action against personal liberties to consume narcotics may seem a small price to pay for Uncle Sam’s approval and investment. The influence that the United States has used to propagate this type of policy around the world is pretty much indisputable and these regression results certainly seem to agree. Obviously (and as always), more research is needed. By my next blog post I hope to have finished most of the statistical modeling and be ready to start work on the final product of this research, which is an article-length paper.