Traces of Cash
L’argent, the 1983 film by Robert Bresson, tells the story of one counterfeit bank note as it is put into circulation and passed from person to person, with disastrous consequences. The effects of this one dishonest act are transmitted to everyone who comes in contact with the bill, until delivery man Yvon Targe tries to spend the money and is arrested, triggering a chain of events that leave him destitute and imprisoned for life. L’argent is a story about the destructive properties of money, but it also a story of invisible traces – a God’s eye study of a virulent effect as it migrates through a community. Bresson uses the materiality of money – the peeling off and counting and exchange of cash – as a motif throughout. The viewer is in a position to follow the journey of the note as it travels from hand to hand, to learn its origins and transformations along the way. Yvon, on the other hand, experiences the malignant effect that has been put into circulation, but is entirely without recourse, because the history of the counterfeit bill is unknown to him. At the climax of the film, Yvon cries out: “Where’s the money?” He never gets an answer.
Today, cash, cheques and electronic payments all circulate as currency, but cash alone does not record its circulation in any way that is legible. This isn’t to say that cash carries no traces of its journey, but that these traces are secret, invisible, and don’t often come to light. Certainly cash carries all kinds of sediment, from bacteria and foreign organisms to illicit drugs and revolutionary messages. A paper note, for example, can carry the live flu virus for up to seventeen days.  Along with microbes and foreign contaminants, studies show that the vast majority of cash also carries trace amounts of illegal substances it has come into contact with; around 80% of British Banknotes contain traces of drugs, with this percentage rising significantly in urban areas.
Money, because of its unique position as a kind of “skin of the state”, is also a conduit for civic dissent. Along with the visible emblems of the nation state, the serial numbers and hidden visual codes designed to prevent counterfeit reproduction, other secret messages sometimes make their way into money. Artist Cildo Mereilles’ Insertion into Ideological Circuits Two: Banknote Project (1970) explored the relationship between monetary circuits and collective ideology. At the time Meireles produced this project, Brazil was experiencing a particularly oppressive period in its military dictatorship. The artist attempted to disrupt ideological circuits by inserting subversive anti-imperialist and anti-capitalist messages into Brazilian bank notes, such as the slogan ‘Yankees Go Home!’, before returning the money to circulation.
Once put into circulation on the street, cash, while retaining its origins in a publicly mandated system, enters another system of peer-to-peer exchange. There is a tension here between the control and issuance of a technology by the state and the ability of the individual to use this public medium as a channel for radical or anti-state communications. Furthermore, because these subversive messages are anonymous, they are relatively safe to put into place. Here cash functions as an analogue point-to-point medium, something that everybody uses but nobody completely controls. Even with the availability of more sophisticated civic media such as Twitter, such practices continue today. In Ireland, where a controversial system known as ‘Direct Provision’ places asylum seekers in enclosures for years on end as they await the outcome of their applications, Euro notes sometimes carry the slogan “No Borders: End Direct Provision”. Other messages might just be humorous, as was the case recently in Canada, where Canadians paid tribute in currency doodles to the late Leonard Nimoy by transforming the face of 7th Prime Minister Sir Wilfrid Laurier on the five dollar bill to that of Dr. Spock.
We sometimes see attempts to trace cash in law enforcement. For the purposes of everyday transactions all bills are fungible and one Canadian $5 is no different from another if you’re using it to buy a sandwich. But for the purposes of security and identification, all notes are not equal; each has a unique serial number that can be recorded and used to identify and trace money in the event of a recall or crime – hence the concept of ‘dirty money’, where particular notes are tainted by misdeed and the practice of ‘money laundering’, where bills are transferred or otherwise shuffled around in order to divest them of their filthy traces. A civic interest in the broader patterns of monetary circulation has led to projects such as Where’s George?, a website that tracks the natural and geographic circulation of American paper money. Similar projects exist around the world, such as Where’s Willy? for Canada and Eurobilltracker for Europe, among others. It works like this: bills are manually entered into the database by a user who lists the unique serial number of the note and their zip codes on the website, marks the note with the website link and finally spends the money back into circulation. The hope is that future users will see the tag, look up the registered note online and add further information about its current location to the database.
By tracking these bills and their migration, a whole layer of metadata about where a token has travelled is brought to light. Scientists have used data from the currency-tracking site to visualise the natural trajectories of paper money and find out what these migratory patterns can tell us, about human nature, about disease and about mobility. Using this data, theoretical physicist Dirk Brockmann discovered that money in the US generally circulates in quite local arcs that resemble, but significantly do not reproduce, state boundaries. The physicist scraped the Where’s George? website for data about how the dollar bills travelled, and used network theory to draw the lines that the cash was unlikely to cross. These borders come to represent a monetary topography that overlays official state lines. In places these monetary lines faithfully followed state borders, but not always; Missouri was divided into Eastern and Western territories, as was Pennsylvania. The ‘Chicago catchment area’, as Brockmann refers to it, also includes a significant chunk of both Indiana and Wisconsin. In taking money as a totem for human mobility, the map illustrates how effective communities don’t necessarily observe state lines. More recently, data from sites such as Where’s George? and Eurobilltracker have also been used to model and understand the spread of epidemics and disease. Because cash is hand-to-hand, it’s still an effective representation of physical connections and proximity, providing a convenient way to study the spread of memes and viruses through a population.
The Data of Payments
Projects like Where’s George? and Ideological Circuits point to the latent patterns and practices embedded in the circulation of cash, ‘un-mined’ data about human mobility, sociality, community and commerce. The buried treasure is really there, but hard to spot. It takes something more for these caulked and disjointed traces to come to the surface, to become information.
Even with Where’s George? or Where’s Willy?, there’s still a limit to how much information can actually be gleaned from a marked bill. Where’s George? asks its users to tag their bill with their zip code, illustrating where the bill has travelled to since it was first registered on the website. From this data you can’t tell what a bill is being spent on, or who specifically passed it into circulation – just where it has migrated to and roughly when. The ‘information’ to be gleaned is impersonal and oblique. But even more crucially, a user has to make the decision to ‘opt in’ to the Where’s George?; a marked bill might interest her enough to log on to the website and register a hit but it doesn’t compel her to do so.
Digital payments change all of this. Instead of cold, hard cash, much of our money now rides the rails of information and communications technologies, over computers, tablets, smart phones and maybe even smart watches. Alongside credit card companies, which have been around since the fifties, mobile network operators, Internet service providers and social media platforms are moving from the transfer of flows of information to the transfer of flows of value, as can be witnessed in the range of different mobile and e-payments options available today, from Apple Pay and Google Wallet through to Tweets for Cash, Facebook and peer-to-peer services like Venmo and Snap Cash. Physical tokens melt into air, to be replaced by SMS texts, cryptographic hash functions or 140 character strings. The result is that most of our transactions now occur in heavily mediated spaces, where data about our transactions is automatically registered alongside a range of other personal data including our identity, location and our social networks.
Internet companies make money by building channels for cash on top of those already put in place by banks and credit card companies, allowing for the secure transfer of payments online, over mobile phones or through well-known social media platforms. The revenue these companies make is generally a ‘toll’ for electronically transferring value from place to place in a secure fashion and/or providing the appearance of synchronous settlement. Payment-related fees are a primary value proposition; but it turns out that these fees may not be large enough to justify these new services. First of all, there is already competition from existing banks and credit card companies in the e-payments space who also rely on similar fees for revenue, and second, there are growing regulations in the US and the EU that limit the amount of fees that a company can charge for monetary transactions or remittances. Another form of revenue then is based not on tolls but on leveraging the value of data that is gathered by the network throughout the course of an electronic transaction. In other words, instead of selling a right of way, throw your roads wide open and sell a little bit of whatever falls off the truck – purchasing information, demographic information – the various bi-products of transacting online.
Data is more valuable than fee-based revenues. In fact, Information about a monetary exchange may have a comparable or greater exchange value than the monetary token engaged in the transaction itself. As Alex Rampell argues, “as society becomes ‘cashless’, companies have a larger business – and a more valuable one – in closing the loop for offline transactions and helping deliver customers to advertisers.” And when data about a financial transaction is taken together with linked data such as identity, location, behaviour and social connections, this yields all kinds of future value propositions. (Or at the very least many of the companies operating on this model are speculating it will). (Meta)data is the new money.
Metadata is the new money
Data can be automatically gathered through the use of a platform, a handset or through a customer loyalty card, through purchase histories, basket analysis and click streams. Mostly this happens in online shopping and payments, but it is also being applied to our behaviour in physical retail spaces. Companies are working to bring the same level of data mining to bricks and mortar spaces through the use of beacon technologies that, through mobile phone identification, can track a consumer’s position as well as their purchases – a practice Verizon Wireless has cheerfully dubbed “cookies for the real world”.
Transactional data is often sold to third party advertisers where it is used for targeted advertising and location-based services or to promote cross-selling. Perhaps people who buy donuts also buy Ariana Grande tracks? Or, to gloss Minority Report, ‘Here’s a coupon for that Gap store nearby. How did those button-downs work out for you the last time?’ This data is also used by retailers to streamline in store operations, pricing, logistics and supply chains. Yet another rising trend is the use of transactional data to introduce tailored financial services. Companies are now coming forward who specialise in personalised credit offerings based on mined transactional data. One such company, Branch, focuses on data emerging from the growing use of smart phones in Sub-Saharan Africa. Branch leverages data from the use of services such as the popular mobile money app M-Pesa and Facebook to offer tailored micro-loans to its clients. In short this involves data mining a user’s transactional data and other kinds readily available through their devices such as demographic, psychographic, social network and even biometric data and using this information to infer a user’s potential credit worthiness and design persoanlised lending criteria. In China today Alibaba, the world’s biggest online shopping platform, have released the ‘Sesame Score’, a social credit rating based on transactional data gleaned from the site and from its partners such as the taxi service Didi Kuaidi. The score can be used to procure credit from Alibaba’s financial wing, but users are also encouraged to publicly advertise their scores to their peers. Baihe, one of China’s largest matchmaking services, has even teamed with Sesame to promote clients with desirable credit scores on their websites. Using the Sesame Score as a testbed, the Chinese government recently announced that something similar will be mandatory by 2020. This score will contribute to the outcome of obvious processes such as loan applications but also to Schengen Visa applications, with the launch of Sesame’s Credit Visa system.
While most companies are a little cagey about specifically what kinds of data points among the 5000+ accessed build these real-time credit ratings, payments, purchases, the financial solvency of your relatives and social networks, your online sentiments and even how often you charge your phone get mentioned. Buying too many video games might also be a bad idea. Having wealthy friends might positively impact your credit rating. But having fewer friends might impact your credit score further – who can say exactly?
It’s not that using consumer data to profile populations or segment markets is a new practice, but that these practices are much more extensive than ever before, and more personal. Profiling companies have started to produce individual rather than aggregate profiles by combining data such as credit and store card transactions, click streams, social media posts and other kinds of personal data. Instead of producing a widespread message, advertising companies can now produce fine-grained communications for a small number of targeted consumers. Frank Pasquale has suggested that the precision of such analyses far exceeds anything we could previously have imagined, not taking in market segments such as 18-24 year old female in a particular zip code but hyper-segmented markets as precise as ‘gullible elderly with history of gambling’ or ‘had a daughter who died in a car crash’. Probably the best-known example of this is the story of the chain store Target, who correctly inferred that a teenage customer was pregnant, based on her purchasing history, and inadvertently broke the news to her parents with some thoughtful baby crib flyers.
We are seeing new identity-based monetary systems, which in turn have the power to produce greater inequalities not only in terms of economic wealth and opportunity, but also in terms of citizenship and everyday life. The machinations of these systems produce new algorithmic credit castes and financial instabilities. In sum the effects of such dataveillance are unevenly distributed; as academics such as Helen Nissenbuam, Joseph Turrow Frank Pasquale, and Martha Poon have all explored, such computerised and software-based systems don’t proceed without bias but have a tendency to reproduce and strengthen existing racial, gendered and economic biases.
What’s more, transactional data doesn’t only influence decisions that are made about us in the present. Categorisation and segmentation also play a constitutive role that shapes our futures whether we are aware of it or not, foreclosing certain possibilities and suggesting others. The extreme example of this logic at work might be something like Amazon’s predictive shipping algorithm, patented in January 2014, which allow for anticipatory logistics in response to customer analysis, a kind of ‘what you want before you know you want it’.
But these techniques also have implications beyond being pushed unwanted recommendations or sold products that appear to fit with our algorithmic profile. Secondary data and data derivatives are also repackaged and sold to governments and institutions where they are used in new and unintended ways. Increasingly transactional data are used in security, governance and law enforcement; for example, in the targeting of anti-money laundering practices or tax evasion, as is the case with the Italian Redditometro, an algorithmic income meter designed to compare people’s spending patterns with their income to detect possible tax fraud. Or perhaps your data might be used to infer something about your trustworthiness or to defame your character in some future context – like the LA supermarket that threatened to disclose one of its customers alcohol purchase history in a slip and fall lawsuit. Or more seriously, this data might also be used to profile you as a terrorist or security threat (helpful hint: don’t purchase a pressure cooker and a backpack on the same weekend).
Tactics for Secrecy
All of this throws up ethical considerations about who owns and controls transactional data. Is transactivity a ‘commons’, held and produced by people and outside the remit of the state or the market, or is payment data part of the purview of corporations and government agencies, a quid pro quo for convenience or a necessary ingredient in the prevention of tax fraud or terrorism? And what’s the answer for those who wish to keep their private transactional data private? According to Peter Sunde, founder of the PirateBay, cash is now the last bastion of privacy in a world where all our transactions are subject to dataveillance. Hence the enduring popularity of cash payments in illicit transactions: “If everything is traceable you start thinking about your purchasing behaviour. You need cash for anonymous behaviour”. And yet we can all probably recognise the utility and even the necessity of electronic transactions in a lot of instances.
As a result, communities of artists, hackers, engineers and academics have worked to develop obfuscation tactics to make transactional data more secret. These include tactics that hide money’s footprints in a mesh of false traces, that work to erase the traces left behind by digital money, or, by swapping accounts and tracking objects (SIM cards, loyalty cards) within a group, produce false and confusing data.
A decade ago this included face-to-face loyalty card swap meets, where users collectively pooled their purchases on a single loyalty card, or periodically swapped loyalty cards to maintain the savings benefits while confusing transactional data that was being gathered about them. The greater the number of people involved in a card swap and the further the cards travelled and circulated, the greater the obfuscation of this data. Gradually these practices were performed over social networks and mailing lists such as the now defunct Cardexchange.org, a website which coordinated swaps through mail. In Rob Cockerman’s culture jamming project, The Ultimate Shopper the artist printed and circulated the code on the back of his loyalty card to other users. Another example by Rob Carlson, Giant Bonus Swap Card Meet, created an online card swapping system for the Baltimore and Washington DC area where participants could enter their card numbers into a form on the site and then print out and paste someone’s bar code onto their loyalty card.
Contemporary tactics focus on digital data obfuscation. Applications such as Cachecloak produce obfuscation strategies for Location Based Services. While Cachecloak doesn’t act on transactional information specifically, it acts on other contextual and locational information gathered at the same time, obscuring a user’s location by surrounding it with other paths through the propagation of ambiguous data. Sites such as Trackmenot seek to obfuscate search and transactional data, while Adnasueum engages with the negative outcomes of targeted advertising so a user doesn’t have to. Another approach involves shirking credit cards and accounts in favour of Prepaid options. Prepaid Gift Cards, once exclusively the remit of the under-banked or those without access to credit, are now being strategically used for their anonymity and privacy benefits. Another less cumbersome option is the use of virtual credit card services. Privacy.com, for example, creates a virtual credit card number for every transaction, marketed as a solution both to increased online fraud but also to data mining.
Probably the most significant tactic of all lies in the development of cryptocurrencies – digital currencies and payments systems that use cryptographic techniques to guarantee the security of a monetary transaction. Unlike a publicly mandated payment system, which has some centralised authority such as the Central Bank or the Federal Reserve as its guarantor and trusted intermediary, or commercial payments systems, which rely on centralised platforms and data infrastructures, cryptocurrencies decentralise both the infrastructure of monetary creation and value transfer. Instead of a central trusted authority then, cryptocurrencies use cryptographic techniques to verify and secure transactions. This system not only ensures that no centralised entity or institution is the gateway for monetary supply or transfer, but also affords a high degree of anonymity in digital transactions, impossible when transacting through the gateways of a private or publicly-mandated network. The best example of this is of course Bitcoin.
Bitcoin offers its users some level of anonymity in their transactions, because cryptographic transactions require public and private keys. Private keys are analogous to a password or pin number, not to be disclosed with anyone, while public keys are analogous to a bank account number, with the key difference that these public keys aren’t directly associated with a particular identity. At the same time, because of the decentralised design of the currency, the history of previous Bitcoin transactions are all perfectly legible, detailed in a distributed public ledger (called the blockchain) where anybody who wants to can see them.
The relationship between transparency and personal privacy in Bitcoin is different to other digital payments, therefore. The protocol offers a level of anonymity because Bitcoin addresses aren’t readily associated with a public identity. This makes it difficult to link transactional histories to individual users. Consequently Bitcoin is often used for illicit transactions online. Bitcoin continues to be associated with transactions on the dark web, with darknet marketplaces like Silk Road and its successors, such as Evolution, processing more transactions than Bitpay. However, in another way the Bitcoin protocol is more transparent than other payments systems because the entire history of all transactions on the network is publicly available for all time (or until people get bored of Bitcoin, whichever comes first). Because of the public nature of the information, transaction data on the blockchain and associated lists such as message boards, public forums and social media networks can be mined to identify particular users and patterns in transactions. Indeed, this is now the aim of anti-money laundering units worldwide
More recently, extensions to the Bitcoin protocol have attempted to tackle some of these weak points through additional obfuscation strategies. With Bitcoin laundering, for example, it is possible to use various techniques to cover your tracks when using Bitcoin, such as pooling funds together into a shared wallet with other trusted individuals. Services like Bitlaunder or Bitcoin Mixer remove traces of previous ownership from transactions to scramble transaction histories. Zerocoin have also developed the Zerocash protocol for more truly anonymous payments. This uses added secure cryptographic techniques to ensure that “transactions record neither the payment’s origin, destination or amount”.
Strategies like prepaid cards, privacy plug-ins and cryptocurrencies provide tactics for users who, for whatever reason, don’t want to their traces discovered. Many of these obfuscation strategies also require a degree of technical expertise that is outside the remit of most consumers, or present playful short-term fixes to what is ultimately a more widespread problem. Still too, these practices have come in for criticism for designing the perfect hack for aiding and abetting tax evasion or criminal practices. As an example, the Paris attacks in November 2015 were partly funded by anonymous prepaid debit cards.
New economies of Metadata
According to Marshal McLuhan, “’money talks’, because money is a metaphor, a transfer and a bridge… It is action at a distance both in space and time.” Money is both information and infrastructure, in other words. It includes the symbols that represent an abstract economic value but also the cultural and technical infrastructures that support their circulation: physical tokens, kin communal networks, fibre optic cables and electromagnetic spectrum. Money is the medium and the message, its true. We can now add a third facet to this and say that in the shift to digital payments money is also the hidden traces of its own circulation, what Jeffrey Pomerantz calls administrative metadata, which provides information about the origin and maintenance of an object, and use metadata, which provides information about how an object has been used. These new economies of metadata demand our attention. What does it mean for the future of money when all monetary transactions can be indexed and traced? When the bi-product of exchange now has its own exchange value?
The trends discussed in this article throw open questions about the artefacts of exchange, a valuable data commodity that until recently wasn’t a ‘thing’ or an asset at all, but a hidden trace that didn’t extend much beyond the imaginaries of Bresson’s forged tokens. With the shift to digital payments, and in turn a shift from publicly mandated monetary systems to commercial IT platforms, we see changes not only to how money is issued and guaranteed but also to how cultures of exchange are managed and mediated. Above all is the question of whether transactional data – a bi-product of exchange – should be available for inspection or commodification and whether appeals to security or consumer convenience are really enough to justify these practices. Does providing an infrastructure for exchange provide a license to data collection? And how will the collection and analysis of this data influence questions of financial equality, citizenship and security going forward? And yet other practices discussed in this article hint at how cultures of exchange always contain an element that is excessive to capture, a hidden trace, something in art works, hacktivism or everyday monetary practices that escape circulation in the system. Whether this counts for anything remains to be seen.
Many thanks to Johannes Lenhard for his editorial input and to Quinn Dupont for his helpful comments in the draft stages of this article. Thanks to Nigel Dodd, Where’s George, Cildo Mereiles, Dirk Brockmann for image permissions. Rachel O’Dwyer’s Research is supported by the Irish Research Council.
 Based in turn on the Tolstoy novella The Forged Coupon (1912).
 Scientists at the Institute of Genomics and Integrative Biology found 78 different disease causing microorganisms on a single bank note. http://indianexpress.com/article/india/india-others/your-money-is-dirty-government-study-says-currency-notes-carry-disease/
 Foster, R.J., 1999. In God we trust? The legitimacy of Melanesian currencies. Money and Modernity, pp.214-231.
 Where money sent to a third party appears to instantly appear in their account. In reality companies like PayPal achieve this by fronting the money while waiting for the Banks to settle in the background.
 To use an expression favoured by telecommunications, we can think of these two value propositions – payments fees versus payments data – as the difference between a ‘smart’ and a ‘dumb’ pipe. A dumb pipe, as the name suggests, simply moves information through its channels without any overarching cognisance of what this information is or even where it has originated and previously travelled. A smart pipe, on the other hand, exploits its position as a network carrier to produce other streams of value by aggregating and subsequently monetising the latent data transmitted over its channels. And today the wealth of networks comes from being a ‘smart pipe’.
 Nissenbaum, H. (2001). How computer systems embody values. Computer, 34(3), 120-119.
 Turow, J. (2008). Niche envy: Marketing discrimination in the digital age. MIT Press Books, London.
 Pasquale, F. (2015). The Black Box Society. Cambridge, MA: Harvard University Press, 36, 32.
 Poon, M. (2007). Scorecards as devices for consumer credit: the case of Fair, Isaac & Company Incorporated. The sociological review, 55(s2), 284-306.
 A very good book to look at in this context is Finn Bruton and Helen Nissenbaum’s Obfuscation: A User’s Guide for Privacy and Protest (Cambridge: MIT Press, 2015).
 Bitcoin is a decentralised electronic cash system that uses peer-to-peer networking, digital signatures and cryptographic proof to enable users to conduct transactions without relying on a third party or intermediary to guarantee and monitor the repartition of funds. Nodes broadcast transactions to the networks, which records them in a public ledger known as the blockchain after validating them within a proof of work system – a cryptographic hash function performed by all nodes in the network.
 Marshal Mcluhan, M. Understanding Media: The Extensions of Man (London: MIT Press, 1994), p.136.
 Jeffrey Pomerantz, J. Metadata, (Cambridge: MIT Press, 2015), pp. 17-18.