Thursday, 14 July 2016 13:16

Gun Control and Learning to Make Data-Driven Decisions

Posted by 

We live in an unprecedented age of information and computing power -- yet whether publicly or privately, we still make decisions largely on how we feel rather than based on hard information.


With this massive influx of data and computing power, when it comes to issues like global warming or gun control -- and much of what is being shown as data is false -- we are more likely to decide based on how we feel about an issue than spending a comparatively little time, compared to our ancestors, figuring out the facts and then making what would be the right decision for us.

Instead, there was a counterproductive rebellion in congress. Its only direct impact wasn't to prevent or even reduce the chance of another mass shooting, but rather to result in the sale of thousands of assault weapons. (Why don't we ever talk about how these failed efforts sell tens of thousands of guns?)

From the data I'm seeing, the Democratic Party, not the NRA, is actually responsible for selling more assault weapons. Maybe gun companies are funding the wrong party?I'll share my thoughts on fixing the big problem, which requires shifting from emotion-driven decisions to data-driven decisions, and I'll end with my product of the week: Politifact.

Orlando Shooting

What most troubles me about the Orlando shooting is that it was preventable. Within a very short period of time after the shooting, we knew that the shooter was troubled, that he had anger control issues, that he was violent, that he had been communicating with terrorist organizations, that he had been on a no-fly list (but was no longer), that he'd recently bought several guns and ammunition after being turned away by a gun store, and that he had been behaving suspiciously.

That data should have resulted in an attempt to ascertain whether he was a risk and if so, to do something about it. Those separate data elements are correlated only after a crime, not before, so they are really great on proving guilt and establishing why something was done -- but not preventing it from happening in the first place.

Did we focus on fixing that problem? No, the government separated into two camps: one protecting gun rights and the other attacking them. Two bills resulted and failed, but had they existed prior to the Orlando shooting, they would have had no impact on it. It's as though after hitting the iceberg, the crew on the Titanic argued over whether ships should be allowed to sail into shallow waters.

We often don't know the cause of a problem, but we do in this case, and yet both sides are doing their best to avoid talking about it. To date, there is still no major effort to connect the data elements needed to effectively mitigate mass shootings. So this isn't about prevention -- it is about manipulating an outcome. I expect it has more to do with the fact that the NRA supports right-wing politicians more than actually saving lives.Now, for the bills in Congress; what is funny is that the only amendment that actually might have prevented the Orlando shooting was from Sen. John Cornyn, a Republican from Texas, of all places.

The data we would need to decide on the two core bills blocking gun sales to those on the no-fly list and requiring background checks for gun show buyers -- given neither would have impacted Orlando -- is how many people on no-fly lists commit gun-related crimes, particularly mass shootings. The same is true for those who buy guns at gun shows.

The bills are supposed to correct a problem, but there is no data suggesting a major problem actually exists related to these two behaviors. The first step in making a decision should be to figure out if there is a reason to make it, and given this Congress is kind of famous for not doing stuff, picking things to do that don't have any impact seems a tad, well, wasteful.

I have an idea. The CDC captures deaths by cause. Why doesn't Congress use this list to prioritize funding to keep citizens alive? Right now, heart disease is No. 1, followed very closely by cancer.

Gun deaths take about 2.5 percent of the toll of these two diseases combined. More people die from car accidents, which suggests that fixing the car thing should have the higher priority. I should add that two-thirds of the tracked gun deaths were suicides. In other words, you are twice as likely to shoot yourself than to be shot by a bad guy -- but let's be clear, it is a choice. If two-thirds of the gun problem is folks shooting themselves, then maybe the focus should be on not selling guns to suicidal people.

In the end, particularly given how difficult it is to get gun bans, programs on gun safety and to better help those who are suicidal likely would have a far larger impact than the bans. (Seriously, we weren't able to ban drugs effectively, and it isn't clear gun bans are effective in the U.S.). You have to prove causation -- not gut feeling, but actual causation -- and that is missing from the discussion. To make a decision to change something, we need to know it actually will work.

Avoiding Manipulation

If you are data driven, you become far harder to manipulate. Take global warming. If you look underneath the debate, there is no private interest that has been identified as fabricating it as a hoax. The data sources are largely government agencies and schools. Oil companies and the people who own them are the main creators of the counterarguments.

That one little data tidbit should have you questioning at least the ethics of one side. Had we known that doctors were behind the idea that smoking tobacco causes cancer and that the cigarette companies were behind the challenges to it, we would likely have addressed the issue more aggressively and effectively.

Even if you toss out global warming, its causes -- pollutants -- are directly tied to health problems like cancer, which should top the list of congressional goals. Based on the data, the stronger near-term problems tied to mortality are those tied to pollution, not necessarily global warming, which suggests we should make the necessary fixes anyway.

Wrapping Up: The 3 Elements of a Data-Driven Decision

So here is my recommendation. When considering any major decision, break it down into three data elements:

The first is priority: Is there another decision you should make first? For instance, should you buy a new boat or pay down your mortgage first? If you pick the latter, then you don't need to spend time researching which boat to buy.

Second: what is the problem you are trying to solve? What is the primary reason to buy that car, boat or home? That will define which data you need to make the decision. For instance, if the goal is to have a relaxing vacation because you are overstressed, then you should prioritize your choices by placing exciting but high-stress destinations last. Otherwise, you might do just the opposite and return more stressed than when you left.

Finally: what is the data that defines the decision? In particular, validate the data source to avoid being manipulated by bad data. If you are being shown numbers by a timeshare sales person, maybe it would be good to get relevant data from an unbiased source before making your decision.

Applied to gun control, there are far higher priorities (I'd argue a crazy dictator with nuclear weapons might be a slightly higher priority with regard to problems to fix, for instance).As for the problem to solve, the Orlando mass shooting resulted from the lack of a working early warning system -- not the tool used.

Finally: what is the data that defines the gun decision? What it showcases is that the proposed gun legislation would have no impact.All of this suggests we are just being manipulated, and personally I'd rather Congress actually focused on things that would get jobs done. Given that the bills didn't pass, they aren't even doing a good job of manipulation.

Data is your friend.

There are a lot of interesting websites that provide insight into the information that surrounds us. Some, like Ranker, work kind of like Yelp and rank broad categories like companies and politicians by opinion. The problem is, it isn't objective. Like Yelp, it's likely easily manipulated.


Politifact, which won a Pulitzer Prize, uses actual fact-checking to determine whether a candidate is lying. It is relatively transparent, so you can read how it reached its conclusion and decide for yourself if the analysis is valid.One quick place to look is its Truth-o-Meter, which maps a politician's claim to the related fact. Rankings are "true," "mostly true," "mostly false," "false" and "pants on fire." The last one adds color by implying only idiots would believe a false fact.

There currently are 202 pages of ratings, but they actually make a quick read and can tell you pretty quickly if you are being misled on a major issue or comment, particularly if it would make you look stupid to repeat a particular "fact" at a party.Both Clinton and Trump should read this regularly, because both are quoted as saying a number of things that are pretty stupid (but here Trump's lead is huuuuge -- if he farts, we could lose California).

In any case, because Politifact is one of the major forces trying to help us make decisions on facts -- not emotions -- it is my product of the week. And for godsakes, don't forget about Snopes when it comes to not forwarding something that would make you look stupid.





Read 111 times

Leave a comment


Get Exclusive Research Tips in Your Inbox

Receive Great tips via email, enter your email to Subscribe.
Internet research courses

Contact Info

AIRS Association
5800 Ambler Drive Suite 210 Mississauga,
Toronto, ON, L4W 4J4 Canada
Email: Support[@]airsassociation[dot]org
Telephone: +1(905) 228-2348
Telephone: +1(905) 488-0753
Fax: +1(905) 228-6187

Subscribe to AIRS Newsletter

Receive Great tips via email, enter your email to Subscribe.
Please wait