Posted by The Weakonomist in Wednesday, October 7th 2009

My job function has changed recently and while I’m still not yet in a position to talk about it, I can tell that my new role allows me to meet people from all walks of life.  Not since high school have I been exposed to a wealth of diversity as I have in the last few months.  Only in high school I wasn’t analyzing every single person I see and trying to break them down on a financial level.

My prejudice is both a gift and a curse.  For most instances it is a curse though.  Though I’m not proud of the hypothesis of my observations, I am scientific enough to know what I’m seeing may not be accurate, and to know late these observations take root in my conscience.

The working hypothesis I’ve got is that there is a measurable (negative) correlation between income level and percentage of body covered in tattoos.  If you aren’t into statistics, a negative correlation means that when one goes up, the other goes down.  So as the observations go, the more tattoos you have, the less your income and vice versa.

There’s no need for me to explain what I’m talking about here.  Obviously the guy pictured to the right is an extreme example, but it’s also obvious you’re not going to see this guy making deals on Wall Street and handing out business cards with the letter “MBA” after his name.  I’m not sure who he is, but it is easier to imagine this guy is a “lifer” in San Quinton.

The great thing about what I’ve been noticing is it’s really easy for an academic to test.  Collect 1000 tattooed people and measure their incomes to tattoo correlations.  It’s probably been done, but I don’t have the means to discover this.  For the sake of this post, let’s assume that the hypothesis cannot be disproved; in the statistics world this is as close to a “you’re right” as it’s going to get.  Assuming there is a negative correlation between tattoos and income what conclusions can be drawn from this?

NONE

This is known as correlation without causation.  Just because you observe something doesn’t mean a conclusion should be drawn.  You can’t justifiably say that when you see someone covered in tattoos they probably don’t make a lot of money.  Likewise if you see someone with no tattoos you can’t conclude they do make a lot of money.  Most importantly, you can’t make an assumption that the choice to get “inked” directly resulted in someone’s income being less.

In data analysis you will always observe exceptions to the rule.  There are many tattooed people out there that make more money than you or I ever will.  Professional athletes are the first to come to mind.  Many are covered in tattoos and make millions of dollars a year.  These are known as outliers and exist in almost any analysis.  But they are exceptions only, and in the case of tattoos they are likely not frequent enough to skew the results.  On the other side white-collar America has plenty of tattoos as well.  However most of these are worn in places covered by clothing, therefore the wearer is not interested in showing the world something that is merely more personal to them.  Their contributions to the correlation would also not be significant due to a very small percentage of the body being tattooed.

So what was the point of this post?  You’re probably thinking I’ve wasted your time and were hoping for me to actually cite an article that has information about this.  It’s important to recognize when information or prejudice is beneficial or not.  There is nothing gained from knowing if a correlation exists.  In fact you probably lose a bit of open mindedness about tattooed people you meet for the first time.  They say knowledge is power and for the most part I agree, but knowledge like this is a handicap.  As fascinating as it would be to know if there is a correlation and how strong it might be, it’s better for us not to know and always have an open mind about everyone we meet.