Guns and statistics

A few years ago I started another blog titled "Scientosis," but I didn't keep it up for very long. In that short time, I wrote an article about guns based around a video from a pro-gun advocate to highlight how statistics are used and misused in the gun debate, and how bad logic flourishes because of it. In light of the recent renewal of the gun debate, I have opted to post that blog article again, with some alterations to update and improve it.

 

Mark Twain once said "There are three kinds of lies: lies, damned lies, and statistics."  Twain is someone I greatly admire but this one of the few times a quote of his fails.  It is certainly true that people use statistics to mislead, manipulate, and...well...lie.  But the point is that if you know what to look for and what question is being asked, it is possible to tease out facts from opinion and separate poor from good ideas.  The statistics don't lie, but the people using or misusing them might.  

There is another reason why I want to discuss statistics, and it is due to the issue of polls leading up to the 2016 election. The majority of these polls showed Hilary Clinton (D) with a strong lead for much of the lead-up to Election Day, and her loss caused quite a surprise to those of us who watched the polls. Polls are indications of what a subsample of the population believes at a given point in time, but are not actual predictors of the future. In addition to this, polls of public opinion will also necessarily have some amount of error because these polls are conducted on small sample sizes. Meaning that a 1,000 randomly chosen participants should (if truly random) reflect more or less the sentiments of the population at large, but polls are often not truly random. There are a variety of reasons for this, and most of them are benign. For example, if one conducts a phone survey for a poll, this relies upon the people you call to have a phone and to answer calls from unknown numbers. Internet polls can easily be manipulated by people voting numerous times. And often times the pollsters may select a targeted demographic or region to poll, and that region may not reflect the country as a whole. If you want a good recommendation for a blog/website to get more information about statistics and polls in general, try Nate Silver's fivethirtyeight.com. 


I have pondered how to best present my argument on guns and statistics and I realized it might be best to do it with this video on gun rights.  http://www.youtube.com/watch?feature=player_embedded&v=Ooa98FHuaU0


I'll start off by saying that whatever your opinions and views are on this issue, the point is to analyze what he is saying and how.  You can make up your own mind as whether or not to amend our current set of laws, but using arguments like those he makes are going to put you in a position of being a manipulative and dishonest individual arguing partial truths so as to further an agenda. (I'll add to this as I update the post to say that in the wake of the Parkland, FL school shooting, I don't think we can sit back in good conscience any longer and do nothing. We must start standing up to these bad arguments and let facts inform our decisions.)


So, here are the basic points of what I have found wrong with what he is saying, and I recommend you watch the video before reading my blog post that follows as I will directly address things said therein.


First off, he starts off with points I can agree with. That there are other factors that must be considered, but he digresses quickly.



Comparing 2 years, rather than trends. He compares 1992 and present. Only comparing 2 years worth of data is misleading because crime rates can spike and drop on a year to year basis. What matters is the overall trend over the course of several years. There is a reason no one just looks at 2 years worth of data when making comparisons between other countries and between states.  The same is true for comparing other sets of data like temperature and climate in the climate change debate.  If you only compare averages, and/or maxes and mins for any 2 years of your choosing, you could make an argument for stasis, decreasing temperatures, increasing temperaures or indeed anything you want.  The point is that longer term trends must be evaluated over more than what would equate to 2 data points (2 years worth of data).



Metropolitan areas are high concentration areas of people where crime rates should be expected to be higher.  It is not just gun violence and it is not just the US where this is the case.  Or indeed only modern times.


This is for 2 reasons: First, increased frequency of people interacting with one another means more opportunity for a crime to occur. Basically what I am saying is that when there is more opportunity, there are more occurences of some rare thing. If you only go out into the forest for an hour looking for a rare species, you may never see it. But if you set up a camera and leave it out for a year, your chances increase exponentially.

The second is poverty level and average incomes. Low-income areas are also larger and more populated in Metropolitan areas, and cities have an overall greater disparity between income levels. In rural areas, there are fewer high-income families and fewer low-income families overall. The differences between rural and urban America is also connected to the overall wealth of the areas, and things like the price of basic commodities, such as housing costs. Basically, your money goes farther in rural areas than in metropolitan areas because things are almost always much cheaper there. A $20,000 salary in New York City is very low income and not feasible to live off of because of the price of commodities like gas, milk, bread, transportation, rent, etc. In rural areas (like where I grew up in TN), the price of these commodities is much lower. Getting a cheap car is pretty easy. Wal-Marts are frequent for low cost shopping. So, living off of $20,000 is much more reasonable (rent is ridiculously low in areas like that whereas you may not be able to find an apartment in New York for $20,000 a year)


The "true trouble spots" when looking at a neighborhood by neighborhood basis is also not surprising. Bet it is highest in low income areas. And there are more of those in metropolitan areas. These are ignored by everyone because there is no quick fix for them. It can only come from improvement of those areas from within. One thing these areas need is better access to education. Part of solving that problem means investing in the school systems in these areas, a plan that is contra to what the current Secretary of Education (Betsy DeVos) wants to do. This is why so many people highlight her plans as being inherently racist because her policies would favor taking resources away from schools that are disproportionately diverse compared to the US population at large. 


"Who is working on improving that." That is why people want universal healthcare in the US. Reducing the cost of basic commodities is highly valuable to these communities and improving the quality of the education can have a big impact too.


How do you improve the lives of the lowest income portion of the population? Man that's a tough question. I have already said education twice, but what else could you do? I don't know the answer to that.


Can you even do it in the short-term? Quick-fixes rarely work for any extended period of time, and we need long-term fixes.


Education is key. Improve education overall and you improve the ability for people to access a wider array of stable career opportunities. What people want aren't just good paying jobs, but jobs with stability and benefits. It is not unreasonable to ask corporations (making record profits) to pay their employees enough money to live off of such that the taxpayer isn't the one subsidizing their basic cost of living. To make this clear, I'll give an example: someone who works and is on food stamps is not being paid enough by their employer. What that means is that your tax dollars are going to feeding that person so that they can go to work at a job where they are underpaid. You are subsidizing corporate profits via your taxes so that someone can eat.


I don't know of anyone that has said the US has a lower crime rate than England. He is cherry-picking what he is saying here and building a bit of a straw-man.


The murder rate is still lower in the UK. He seems to want to dismiss this. Isn't that what he is criticizing everyone else of? He is not giving a suggestion as to why their murder rate is lower, but I have a guess I will get to in a second.


Gun crime is lower. I think this is what most people are saying instead of just crime overall.
Yes, the UK has more violent crime, but that may be due to something I have already mentioned, and that is that the UK has fewer metropolitan areas, but it also has a lower population:
Percent of Americans living in Urban and suburban areas = 82%
Percent of UK residents living in urban and suburban areas = 90% http://www.guardian.co.uk/news/datablog/...ing-cities

Crime occurs in metropolitan areas at a higher perceived rate than rural areas (something he said too) and more UK citizens live in metropolitan areas, so it would be expected for crime rates to be higher, regardless of gun laws. But murder rate is lower even though that too should be higher in metropolitan areas. So, the UK seems to have that figured out. Basically, less access to highly efficient murder weapons means lower murder rates. So the total number of metropolitan areas is not what is important, he is cherry picking again.

Crime rates in European nations have dropped too over a similar time frame. That means that the global economy is likely a major contributor, so nothing happening in the US is unique as far as dropping crime rates.


Gun crime has gone up 4-fold since they amended their gun laws, but that is a false positive in a way.  It is true that those numbers have gone up, but they have gone up in the same way you would expect to see crime rates go up because you hire more cops.  There is not really more crime, just more people getting caught.  Think of it like this, you live on a road where the speed limit is and has been 35 mph for as long as anyone can remember, and then one day it is changed to 15 mph.  Are the number of speeding tickets going to increase or decrease?  They will increase because the restictions have been tightened making it easier to be considered in violation of the law, and you have a generation of people who are not comfortable with the recent chances yet.


Another part of the puzzle that needs to be addressed is the frequency of mass shootings. The US is a statistical anomaly there too with a rate higher than other European nations.


Yes, the politicians have an agenda. So do the politicians on the pro-gun side. The point is that it is important to look through that agenda. I don't have cable. I don't watch the pundits. I would have been right on his side pretty much right up to the Newtown incident. Having a kid and being an hour away from something like that will quickly change your perception of such issues.


When has the media been honest about anything anyways? They oversimplify on pretty much every issue because the people watching are not smart enough to understand all of the metrics. What is even more important (perhaps) about modern media is that we are so focused on speed when it comes to reporting, that people often fail to continue to follow news stories to see what may or may not have changed. What is reported first, is not always accurate. If you couple that with the blending of fact-based news reporting and political opinion, you get a television program that conflates opinions with facts. This makes it exceptionally tough for those consuming the information to differentiate between the 2, and that may be nearly impossible for people to do if they rely on a singular media organization for their news.


He cherry picks data at the very end of his own video, committing something he is accusing everyone else of. He uses 2011 only, not the 2000's or some longer period of time. Using one years worth of data is misleading (again). He also only uses rifles in homicides. The point is not that just rifles are a problem.


Solving violent crime? Who is under the delusion that they can solve violent crime? Better education can help, but it won't eliminate it.  He is building a pretty big straw-man here.  I don't know of anyone who is saying that better gun laws will cause violent crime rates to drop, but if murder rates drop and mass shootings drop, that will be a huge step forward.

Basically, he makes all the same mistakes he is accusing everyone else of. None of his arguments actually support less or no gun restriction. The US does not have a crime rate or murder rate that implies more guns equals less crime. So, what exactly is he trying to say? It can't hurt to have this many guns and this few laws? The mass shooting statistics and murder rate say otherwise.



So perhaps Twain would have been better off having said that there are three different ways of lying: saying nothing, telling an untruth, and telling only the partial truth.  This last part is concurrent with statistics in his original quote.