Domesticizing Credit Data: Payday Lenders and the Scraping of Online Information
In making this argument, the paper focuses on a new and poorly understood segment of the high-cost, low value subprime credit market – or the payday lending market, as it is also known. Specifically, it introduces a new frontier in such lenders’ quest for predictive power. This involves a group of well-funded technology startups who are entering the payday lending market and lending at high rates of interest to borrowers who often have either poor credit histories or, in some cases, no credit histories at all. In the UK, the most prominent of such lenders is Wonga.Other notable companies include Kreditech, whose operations stretch across Europe, as well as ZestFinance, which is selling its services to major financial service providers in the USA.
In this new variant of consumer credit lending, we see the exploitation of diverse forms of online data, processed through forms of algorithmic analysis, in the attempt to better predict the repayment behaviour of individuals. This data often appears extremely mundane and to have very little to do with the credit product in hand. Data points can range from a user’s IP address, to the particular browser they use, to their screen resolution, to the plugins they have installed, to information accessed via a user’s Facebook account. In examining this new economic phenomenon the paper presents results from an ongoing collaborative research project that uses cutting edge digital methods to try to ‘track the trackers’.
Specifically, we drawn on a tool called Ghostery. Ghostery is a tracker detector, owned by a company called Evidon, which, according to their own framing, 'shows you the invisible web'. Ghostery works by consulting a 'library of trackers' that Evidon has built up, in part by some of its users having opted to share the trackers detected during their browsing sessions to it. The particular way we make use of the Ghostery plugin has emerged as part of a collaborative project undertaken with the Digital Methods Initiative (DMI) at the University of Amsterdam. The DMI repurposes existing web devices for social and cultural research (Rogers 2013). One of the tools developed by DMI researchers and developers that we have drawn on (and contributed to the ongoing development of) is the 'Tracker Tracker'. The Tracker Tracker repurposes Ghostery’s method of detecting and ordering trackers: it inspects web pages for particular traces of trackers (scripts) and compares them with Ghostery's database. In addition its interface allows the researcher to insert multiple websites (urls) for inspection. The results of this analysis are made available to users in a spreadsheet that facilitates the systematic comparison of the trackers that are found. This can reveal which websites share similar third party trackers and also which companies are main actors within this—usually—invisible ‘fabric of the web’ (Gerlitz and Helmond 2013).
The paper tries to understand the particular tracking technologies that lenders are deploying. One way we do this is through creating tracking profiles of individual website. Through this process we can begin to get a better picture of the density of the data being tracked, as well as to compare data retention policies of the distinct tracking ecologies, what data could potentially be revealed to the user of a particular tracker, and how this might contribute towards to aid credit assessment. Further, in the paper, it becomes possible to understand the potential function of trackers that are unique to one site in comparison to another.
The paper argues that it becomes possible to observe forms of data-mediated but nonetheless domestic monetary interactions, involving the multiplication of the registers of monetary value, with information that surrounds the potential transfer of value, that seemingly has nothing directly to do with the product itself, becoming an asset for the company as seeks to improve its ability to predict repayment behaviour. In the process, we see new asymmetries opening up: such techniques are, as far as it is possible to tell, limited to quite particular segment of the consumer credit market: subprime lending. The fact that such methods are not being used to assess credit in more high status ends of the consumer credit market points us towards the existence of an emergent politics of credit assessment, in which certain forms of borrower submit, more or less unwillingly, to being assessed not just on the basis of their past repayment history, but also their data-mediated online identity. This raises important questions about the politics of transparency and surveillance at play surrounding this particular arena of market-making.