Do banks lose more money to robbery or to stolen pens?

100 Dollar Bill with Pen

100 Dollar Bill with Pen

While making a deposit at the bank the other day and realizing I didn’t have a pen in my pocket, I thought about stealing the pen on the deposit counter. This kind of little theft happens all the time, and we all do it. It’s like doing 51 in a 45 mph zone – we don’t really consider it wrong. So I thought, how much do banks spend in a year in replacing all the pens that all of us miscreants blithely walk away with? Is it more than banks lose to robberies?

Because banks don’t publish their annual pen budget, I have to run this backwards and see where the two amounts meet. I’ve done the research, so let’s do some math. Here are my results.

  • Number of bank branches in US (2010): 98,515. [Source]
  • Cost of a single pen (current): approx. $0.25 [Source]
  • Lost amount of cash and travelers checks, minus the amount recovered in the US (2010): $34,823,034 [Source]

So what I’m really asking here is: ($0.25 cost of a pen) x (?? number of pens) = $34,823,034 amount stolen.

So, we would need to steal 139,292,136 pens a year to equal the amount that banks lose to robberies. Divide that number by  the number of bank branches and the minimum number of days a bank is open every year (365 – weekends and bank holidays = 250) and you get an easily digestible answer. If every bank branch in America loses 5.7 pens per day to theft, then banks spend more on pens than they lose from robberies.

Ta-da! To answer these kinds of questions I was inspired by the book Freakonomics, which I only recently read. Asking novel questions is fun. Thankfully, the internet has made these kinds of knowledge journeys much easier. Imagine trying to do this 15 years ago…


  1. I can imagine doing this 15 years ago…
    “Hello? Can I speak to whomever is in charge of office supplies for your branch? Yes, I’m doing some research on office supply usage in the region and I’m just curious what your annual pen budget is? Yes? You’ve been very helpful! Thank you so much!”

    In addition, I would venture that getting the information directly from the person in charge would be far more accurate than approximately 25 cents apiece, which no office manager in their right mind would pay for bulk pen purchases. The office I used to work at purchased pens at $50 for 3,000 (or around 1.6 cents per pen), and that was for a small business making very few orders. I can imagine a bank would be able to leverage their much larger orders, not to mention the fact that they’ve probably invested in the pen manufacturer, to get an even better deal than that.

    I’m not trying to pick bones here, just pointing out that while the internet has made it easier to find some information, it has drastically hampered the credibility and accuracy of that information, precisely because its ease of access leads to hastily jumped conclusions. In this particular case I’d venture that it hasn’t really saved much time at all, and has led to highly inaccurate results.

  2. 1) A sample of 1 out of a population of 98,515 is insufficient. Some bank branches are very large, some are very small. A sample size of 0.00001% does not control for such factors.

    2) I chose a pen based on the most common pens I have seen in bank lobbies. Click pens, plastic, customized with the name and logo of the bank, bought in quantities of 1,000. They were inexpensive compared to the other pens available that I looked for.

    5) I don’t answer the question, I only point to the fence on which the answer would fall to one or the other.

    I can find as many problems with your research methods as you found with mine. The difference of course is that I actually did my research. I’ve displayed my sources and shown my work, so the problems caused by the one variable that could vary widely should be readily inferred. This was for fun, not a grade. You however, maintain your status as the High Priest of Mystic Skepticism.

Leave a Reply

%d bloggers like this: