An update on our study: new variables for our model
I know its been some time since our team's last update, so I wanted to let you know how things are going. This research is still moving forward - and our preliminary work, has taken us into some thought-provoking new directions before we can arrive at a result.
We have realized that one challenge in estimating effects of new gun policies (including background checks before purchase) is that they will only impact 'new' purchases.
Unless the existing stock of guns is accounted for, results of new restrictions on gun violence will likely be biased towards zero. However, one data challenge for us, is that there are no readily available records of how many guns there are per state per year.
Hence, this research team has spent several months debating and considering proxy variables. We have considered the sale of ammo (digging deeper showed that the data was unreliable) and subscriptions to gun magazines (this has been used as instrument in a very famous past paper --- however, it is less reliable in the era of the internet!).
We are leaning towards using gun ranges per capita as a proxy variable and doctoral students on the team are collecting the data.
With these new variables accounted for in our model, we'll be able to move on in answering the original questions posed in the hypothesis of this project.
As always, please don't hesitate to list any questions/comments below in the comments. We promise to try to answer them as time permits!