Dumb Arguments About Guns is a series devoted to examining the (il)logic of the modern debate over gun rights, gun laws, gun ownership and gun possession. The purpose of the series is to help people focus their arguments and avoid falling into the many pitfalls of the gun debate in this country. In the interest of full disclosure, my own position is that current gun laws are far too permissive. I favor greater restrictions (but not elimination) on the types of guns private citizens should be allowed to own, carry and use.
I ask that people remain polite and on topic in the comments of this diary. Please ignore overly dickish and/or excessively repetitive comments, we can only be derailed if we let ourselves get derailed.
Previous Diaries in this Series:
Dumb Arguments About Guns 1: The 2nd Amendement
If gun control was effective, why are so many more people killed in New York City than Vermont?Throughout the gun control debates are persistent uses of comparisons to demonstrate the value or lack of value of restrictions on gun ownership, gun possession and gun use. In some cases, these comparisons are extremely useful and informative. In far too many cases, however, the comparisons simply don't work. Here I want to talk about the latter cases, though an understanding of why some comparisons are bad provides importance guidance for understanding why some are good. That said, before talking about comparisons specifically, we need to talk a little bit about social science and social science research design.
Japan has strict gun control laws and among the lowest rates rates of gun violence in the world--if we had similar laws in the US, we would have similarly low rates of gun violence.
Social scientists are the folks who study social phenomena (e.g., marriage patterns, economics, government, wealth disparity, etc.). Despite the word science in "social science," not all social scientists are scientists--some are, some aren't and many more are somewhere in between. While it is a complicated story, for our purposes we can divide social scientists into two groups, quantitative and qualitative. In general, the quantitative social scientists are more sciency and the qualitative social scientists less sciency. The quantitative researchers, not surprisingly, focus on statistical analyses. The qualitative researchers spend more time interviewing, participating and learning about people using more impressionists techniques. Both research methods have their strengths and weaknesses. Quantitative methods are better for understanding the broad social patterns of large groups of people, while qualitative methods are better for understanding the nuances of people's lives. For example, a quantitative approach can illuminate rates of gun violence, and the frequency of clinically defined depression in the survivors of gun violence. Qualitative research, on the other hand, can examine the nuanced ways that different people in a community react to gun violence, the marriages wrecked, the friendships destroyed and created, the fears that keep people up at night. Good social scientific research combines quantitative and qualitative methods, using the strengths of one to balance out the weaknesses of the other.
Whether a quantitative or qualitative social scientist, many of the arguments are based upon comparisons. For example, a researcher studying the effects of gun violence on the families of victims would also need to know what families who have not experienced gun violence are like. This can be done explicitly through the use of surveys and statistical analysis, or it might be done by visiting and talking with families in the same community who have not had direct experience with gun violence. In either case, social scientists are constantly thinking about comparisons, what makes a good comparison and what makes a bad comparison--though, it would be fair to say that the explicit methods of quantitative analyses require more explicit consideration of the topic.
Whatever the comparison being examined, the key is to try to compare two groups of people who are as similar as possible in all details except that element that explains the difference being studied. In a sense, this is similar to the methodology used in chemistry and physics labs, where scientists recreate a simplified version of the natural world, alter individual conditions, and see what changes (e.g, with increased heat, water becomes steam: in the absence of heat, it stays liquid). The problem that social scientists face is that we cannot control the social environment, we cannot completely isolate the critical element that we are studying and we cannot alter the conditions to see what happens (in the past, some social scientists have done some truly horrible experiments on living people. These sorts of studies, for the most part, are no longer allowed. See Tuskegee Syphilis Experiment.) Instead, social scientists need to seek out cases where these differences occurred 'naturally,' though obviously social scientists are not suggesting that shootings are 'natural.'
So, the key to a good comparison is that the two cases are as similar as possible, except for the element(s) being studied. A bad comparison, in contrast, would have numerous differences between the two cases in addition to the elements being studied. While spiffy statistics and social theory can help reduce the problems of less-than-good comparisons to some degree, any study that begins with a bad comparison will almost always tend to be wrong. That said, social scientists have developed some general practices that can greatly increase the value and reliability of a comparison--these generally either involve narrowing or expanding the comparison.
Narrowing a comparison involves reducing the size of the two cases being compared. So, if we want to study the effects of gun laws in a small town in MA, we might compare them with the laws in another, nearby town in VT with different gun laws. We can then see what impact these laws have on the prevalence of gun violence, understandings of gun violence and fear of gun violence. Narrow comparisons often lend themselves to more qualitative analyses, for the simple reason that its easier for an individual social scientist to meet and interact with the smaller number of people. On the down side, narrow studies may not provide much guidance for understanding gun violence in places outside the immediate vicinity of where they are conducted. Does understanding gun violence in small towns in VT and MA really help to understand gun violence in Phoenix, AZ?
Broadening a comparison is usually more quantitative. The idea behind it is that all those differences between local communities get cancelled out in the mass of numbers. The weird specifics of gun views in Topeka, KS, Buffalo, NY and Singapore are all diminished by the sheer quantity of views added to the analysis. That said, using a ton of bad data is not necessarily any better than using a little bit of bad data. The problem of comparability is still critical. Thus, rather than comparing rates of gun violence in the US to all other countries, it likely makes sense to compare gun violence in the US to other industrialized countries only. Better yet, we might limit the comparison to industrialized countries with the most similar histories, rural/urban dynamics and demographics (e.g., Canada and Australia).
So, what does all this mean about the gun debates in the US. Simply put, whenever you see a comparison of gun violence in New York City and Vermont, you should get worried. There are simply too many other elements in play to credit gun laws as being the cause of differences between the prevalence of gun violence. That is, does it have something to do with hunting culture, ethnic diversity, urbanism, gangs or something none of us have thought of? The same goes for comparisons between the US and Japan. The countries have wildly different cultures, demographics, histories, etc., etc., etc.
Overall, the key thing to remember is that comparisons are hard to do well. In order to have a good comparison, you need to do a lot of work, a lot of thinking and a lot of analysis. This is true whether you support or oppose stricter gun laws in the US. As the site continues to debate and discuss gun laws, lets all try to avoid presenting--or accepting--weak comparisons in our arguments.
A bad comparison can only lead to bad arguments and bad conclusions.