The circulation of currency has long intrigued humankind. From ancient times, people have no doubt wondered about the history of the coins and bills found in their possession. Where have they traveled? What stories could they tell?

From an economic point of view, money is something of an enigma. Its buying power varies according to the most abstract economic processes. Only some money circulates as currency, the majority exists solely as representations on accounting ledgers. In his essay, “Money, Time, Space and the City,” geographer David Harvey describes money’s this way:

Money is simultaneously everything and nothing, everywhere but nowhere in particular, a means that poses as an end, the profoundest and most complete of all centralizing forces in a society where it facilitates the greatest dispersion, a representation that appears quite divorced from whatever it is supposed to represent. It is a real or concrete abstraction that exists external to us and exercises real power over us.

Even the precise amount of hard currency eludes precise calculation. The Treasury Department’s own “U.S. Currency and Coin Outstanding and in Circulation” report (issued since 1888) that calculates the value of all currency in circulation admits the estimate includes “some old and current rare issues that do not circulate or that may do so to a limited extent.” Who knows how many bills are burned for drama’s sake, or pennies are dropped into wells? Nonetheless, as of June 2008 the government estimates some $826,313,890,478 in total currency and coin is in circulation, or $2,712 per American.

Given its abstract nature, understanding the movement of currency seems beyond any study. How could one track a note with any regularity? What parts of the country have a more dynamic cash economy? Even if you could find out such a thing, would it have any real importance? Is it, as Harvey says, both “everything and nothing”?

The goal of following the movements of individual bills used to be impossible. That was before the invention of the Internet. Since 1998, a growing group (over 4.8 million at last count) of people have systematically marked and logged the movements of 140,485,942 unique bills worth a mere $766,394,212.

This vast tracking project was made possible by WheresGeorge.com founder Hank Eskin. Using his website, hobbyists would write or stamp (marking bills is not illegal, defacing them is) on mostly small denomination bills something like “Track This Bill on wheresgeorge.com.” Recipients of the bill who so desired could visit the website, enter the unique serial number and series, and log their location in terms of the Zip Code. Registered users are notified when a bill they’ve entered (“hit” in the terminology of the site) is entered by someone else.

What can we learn from such an exercise? First, the realities of paper money is running against such an experiment. The Bureau of Engraving and Printing estimates the life span of a $1 bill to be a mere 21 months, and the Bureau printed over 38 million notes per day in FY2007. Such sober statistics may help explain why the most popular bills in the entire system have only been entered around 12 times. In fact, the most popular bill in the system was entered in March 2002 and noted to be in “preety good condition.” The user logging the bill during its most recent entry, three years later in March 2005, noted it looked old. The website has also inspired a passionate following, at least among a tiny proportion of its users. All of the most active 50 users have entered over 100,000 bills, and the most active, “Wattsburg Gary,” has entered over 1 million. The owner of a gun shop in Wattsburg, Pennsylvania’s, Gary’s site profile is a somewhat self-deprecating monument of his dedication to the hobby.

Despite the methodological challenges of relying on such a serendipitous dataset, one group of researchers used it to theorize about contagious disease, but no other uses are easily found online. Let’s explore a bit about what Where’s George can tell us.

The website’s list of most active Zip Codes is thrown off by the location of some of its most active users — Wattsburg, PA unsurprisingly tops the list. Hits, on the other hand, will be less skewed by these super-users. After all, even if they enter a huge number of bills, there’s no way to attract a disproportionate share of marked currency through the money supply. Of course, this could be thrown off by users sharing marked bills, such as I do with my girlfriend’s mother (another user), or happens at Where’s George meetups. Nonetheless, here is the all-time county hit map:

Comparing that with this population density map raises more questions than it answers:

Given the different character of the category breaks, it’s difficult to draw a comparison. The site does have a “Zip Code Lookup” tool allowing you to query how many bills have been entered in each U.S. Zip Code. Using a random number generator, I generated random 5 digit numbers until I had 15 actual Zip Codes (not every 5 digit number is a Zip Code). I then looked up the Census 2000 population and the number of hits in each Zip Code, and did a simple linear trend line using Excel. The results showed considerable variation. The R^2 of the best fit line was a mere 0.44. I decided to plot a few Zip Codes where I had lived. Lapeer, Michigan and Cumberland, Maine, the hometowns of my girlfriend and I respectively, generally followed the pattern. Where I live now, Cambridge, MA, was significantly higher. The Zip Code containing the bulk of the undergraduate students at the University of Michigan is far off the chart.

What if we could analyze the complete set of data against a range of variables. Would we learn users tend to be young, or educated? Would it instead trend towards a Bordom Index, including both college students and rural gun shop owners?

Measuring the velocity and prevalence of cash have other, less obvious uses. After all, one standard economics text notes, “One factor that undoubtedly has affected the currency-deposit ratio is the growth of the underground economy,” pointing out it includes both unreported waiter’s tips and the drug trade, and advising managers of cash to keep an eye on drug policy. Could some sort of sophisticated analysis tease out this sort of information from the master database? Only time will tell.

In the meantime, have you seen any Where’s George bills lately?

> WheresGeorge.com
> Wikipedia: “Where’s George”