Researcher Wang-Sheng Lee talks about Melbourne University gun buy-back paper
ABC Illawarra Breakfast, 23/09/2008 06.51am
Interview with Wang-Sheng Lee, researcher, University of Melbourne, to talk about The Australian Firearms Buyback and Its Effect on Gun Deaths report. Lee has found that the buy-back did not result in any change to gun deaths. Lee says that the research uses the number of firearm homicides from 1915-2004.
Peter Riley: Tuesday morning. How often have you thought back to when the legislation was changed to suspend purchase of firearms and the changes that were put in place back in the ‘90s? Believe it or not, the Australian firearms buy-back and its effect on gun deaths has been the topic of some research that’s been done by a number of people in academia.
The latest bunch of research has been done by Dr Wang-Sheng Lee and Dr Sandy Suardi at the University of Melbourne, the Melbourne Institute of Applied Economics and Social Research. And the findings they’ve found are different to findings that have been found before, if that makes sense. In the past, there was an indication that the research backed up the stand taken by the Government at the time to restrict the amount of access to firearms and that that would actually have an impact on people dying from overuse of firearms. Well, the research has been conducted. And Dr Lee has - has found that it’s not the same, that he’s actually got the data and analysed the data and said, well as far as, gone back as far as World War I, and they’ve got no vested interest in their findings or any affiliations with previous authors on the subject. But found that the research does not indicate any change. At the time of the buy-back, there’s been no change.
So I actually thought this was most interesting and not so much from the shooting perspective, in terms of access to weapons or making use of those weapons, but the data and how they’ve been able to research the data and find a different result. So I sat down with Dr Wang-Sheng Lee and asked him if he can tell me about the data and what he actually did with it.
Wang-Sheng Lee: The data that we used, like one of the time series that we use is like the number of firearm homicides from 1915 to 2004. So, for example, in 1996, there were 104 deaths caused by firearms, right, including the 35 persons in Port Arthur. So we have this entire time series from 1915 to 2004 for the whole of Australia. So what we did, essentially, is to look at this trend in the data, in the time series data, from 1915 to 2004, and check to see if there was some kind of a significant break in the trend around 1996, around the time the law was introduced.
But it’s not a simple like eyeballing exercise that anyone could do, just by looking at the data. What we did, using econometric techniques, was to first find the best way to model the data, assuming that the break did not take place. That means we sort of included some kind of a trend - well in technical terms, what we call like a trend term. And also take into account the fact that some of the past events might affect the current events. And so we modelled - first we tried to model that kind of time - the time series structure in the data. And it’s only after that then we introduced this test for an unknown structural break, where we tried to find - where we pretend we didn’t know that the law existed. We want to let the data speak to us and see if any kind of a break existed around 1996. And we found breaks in the data. But nowhere close to the date in 1996. For example, we found a break in like 1951 but that’s not related to like the Act obviously.
Peter Riley: Is it possible that there would be a time lag between the time the legislation was enacted and a break following that?
Wang-Sheng Lee: Exactly that. That is definitely a possibility. And that is the reason why we use a test for an unknown break. We don’t want to say the break has to occur in ’96 or ’97. Using our test, it’s possible that the break would even occur in ’99 or 2000.
Peter Riley: So did you identify any break?
Wang-Sheng Lee: No, we didn’t then. We didn’t find any break after 1996.
Peter Riley: Amazing. Is it possible that your system of analysing this data is flawed?
Wang-Sheng Lee: We believe…
Peter Riley: I don’t mean that in a nasty way.
Wang-Sheng Lee: Yeah, yeah, no. We believe in our method. We believe in the methodology. After all, the methodology we adopt is based on data that was published in The Review of Economics and Statistics, which is a very prestigious economics journal.
But the weakness in our application, as I’ve mentioned earlier, might be that the data series we have is limited, following, you know, the follow-up data we have from 1996 is limited to eight years. So, you know, we want to go further if possible. For example, if it was 2020 right now, we could redo this exercise and maybe, you know, our results would be different.
Peter Riley: What would be your final conclusion in your paper?
Wang-Sheng Lee: My final conclusion would be that we’re essentially interested in evaluating a particular policy, the gun buy-back policy. It doesn’t seem like the evidence suggests that it made any difference. But we’re not trying to make any kind of overall conclusion that, you know, gun control is good or no good. We’re just evaluating this particular specific policy and it seems that the data suggests that it didn’t make a difference.
Peter Riley: It didn’t make a difference.
Wang-Sheng Lee: Yes, that’s correct.
Peter Riley: Can you see how that might be used by organisations like Sporting Shooters’ Association?
Wang-Sheng Lee: Yes. I see how, you know, they might want to use our results to support their aims and objectives. But I’d just like to reiterate that we’re only evaluating this very particular policy and not making a statement about gun control.
Peter Riley: Interesting. That was Dr Wang-Sheng Lee, who’s from the University of Melbourne’s Institute of Applied Economic and Social Research. Their social researchers have done the research and they found that the introduction of the Australian firearms buy-back didn’t have an effect on gun deaths after its introduction. And they’ve taken new information - a new way of looking at unknown variables and trying to map that. I found that, the way that social researchers can take data and analyse it, most interesting. And maybe you got something out of that.