Smart corruption: Satirical strategies for gaming accountability

Abstract

Although new forms of data can be used to hold power to account, they also grant the powerful new resources to game accountability. We dub the latter behavior “smart corruption.” The concept highlights the possibility of appropriating algorithms, infrastructures, and data publics to accumulate benefits and obscure responsibility while leaning into the positive associations of transparency. Unlike conventional forms of corruption, smart corruption is disguised as progressive, and is thus difficult to spot or analyze through existing legal or ethical frameworks. To illustrate, we outline a satirical strategy for gaming accountability. Identifying the particular mechanisms and outcomes of transgressive activities carried out under the veneer of data-driven transparency, as well as the key actors and organizations most active in gaming accountability, is an important research and political project.

Introduction

Data-driven governance positions transparency at the heart of democratic accountability. In conventional wisdom, information is the antidote to corruption, a way to hold power to account. Such a framing of data as solution is appealing, in part because it sidesteps traditional political divides of left and right and makes a case not for bigger or smaller, but for “smarter,” government. Yet, critical observers note that although technologies can produce positive and emancipatory changes, they can also contribute to accumulation, violence, and inequality by reproducing unconscious and historic biases and creating new ones (Benjamin, 2019; Noble, 2018). New technologies of visibility aimed at openness can paradoxically produce new forms of obfuscation that erode trust, in what Strathern (2000) calls, “the tyranny of transparency.” Big data and its associated algorithms, infrastructures, and public are tools for improvement, but the assumption that more transparency leads to more accountability does not always hold (Fox, 2007; Goldstein and Faxon, 2020; Hetherington, 2011). In fact, such technologies also provide new resources for scamming, fakery, and fraud (Poster, 2022). Particularly in liberal democracies, digital technologies can lay the foundation for what Ang (2020) calls elite corruption, in which new forms of graft, theft, and kickbacks are normalized as a part of public service delivery.

Any discussion of smart governance must consider the alternative: “smart corruption.” We conceptualize corruption broadly as a catch-all term for a wide variety of transgressive activities from actors who range from crony capitalists to unscrupulous bureaucrats (Muir and Gupta, 2018). If the product of using data-driven tools to advance equity and efficiency is considered smart governance, the effect of using such tools to promote self-interest is smart corruption. Unlike surveillance, misinformation, political targeting, and cyber-terrorism, smart corruption is carried out by actors who represent themselves as being held to account by the less powerful while fortifying their positions at the expense of the same stakeholders. Because it is disguised as progressive, smart corruption is difficult to spot. In this commentary, we use a subversive writing style to develop the concept of smart corruption. Our satirical strategy calls attention to some of the many ways incumbents use data-driven tools for their own benefit, representing themselves as vanguards of democracy while cramping capacities for critique.

Smart corruption is not an inadvertent side-effect of a blind faith in the benefits of technology, but rather the outcome of attempts to game accountability. Our analytical approach is rooted in the field of Science and Technology Studies (STS), which understands technologies not as products of linear processes of research and development but rather as contingent outcomes of negotiations, practices, and power relations (Bijker et al., 2012; Jasanoff, 2004). As an analytical concept, gaming is a set of reactive practices through which actors adapt to new technologies, rules, and ideologies in ways that shore up their power and profit (Espeland and Sauder, 2007). It is the process of working the system to one's own advantage (Cotter, 2019). Applied to smart corruption, gaming speaks to the possibilities of appropriating databases, algorithms, sensors, and other forms of information technologies for personal or organizational gains, undermining systems for accountability while aligning with the positive associations of transparency. The combined fustian rhetoric around smart systems and their highly technical nature provide the ideal opportunities for powerful actors to adopt the “aesthetic practices” of transparency in order to signal emancipation (Ratner and Ruppert, 2019) while simultaneously engaging in exploitation. Like data itself, gaming is not inherently “good” or “bad,” and distinguishing between emancipatory and exploitative gaming will invariably involve substantial “ethical work” (Ziewitz, 2019). However, we believe that naming and shaming certain practices as exploitative is an important, though challenging, analytical and normative goal.

Below, we present a strategy for smart corruption, written satirically with the likes of a smarmy politician or profit-hungry corporation in mind. Our playful tone is meant to invoke a management consultancy presentation, and, in so doing, provoke reflection about the dominant narratives around digital technology. In using satire, we revive a long tradition in STS that blends analysis and humor toward theoretical and practical goals, and by no means to make light of a grave issue (Lynch, 2009). In the final section, we shift tone to review foundational theoretical insights, offer new avenues for empirical research, and suggest ways to foster more just technological futures.

Gearing up to game accountability

Data are transforming how we govern and do business. Whether a bold leader dogged by the persistent need to seek public approval or a middle manager worried about who's watching, data isunwinding old rules of the game. The speed of change attests to the power of data-driven technologies. Left unchecked, they threaten operating models. But digital transformation is also rich in opportunity. The information age requires new strategies that harness the momentum around tech-savvy transparency to streamline operations, offset risk, catalyze growth, and accelerate the shift to a win–win world. Here, we show you how to leverage insights from expert research on three arenas—algorithms, infrastructures, and data publics—to maximize the true potential of data for your needs (Figure 1).

Figure 1.

Gearing up to game accountability.

Pro-bias algorithms

Algorithms can be used to realize your goals far more consistently, predictably, and cost-effectively than any human organization. At first, automation seems to wrestle authority away from rank-and-file bureaucrats and into software and all sorts of mind-numbing e-governance platforms. But, with thoughtful design, machine learning can relieve management of decision fatigue while optimizing performance. The key lies in seizing upon perceived neutrality while leveraging algorithmic biases. Remember: one person's algorithmic harm is another's algorithmic gain.

To optimize algorithmic gain, streamline algorithmic bias with your strategic needs. Although earlier generations of management techniques focused on identifying and removing subconscious preferences from human decision-making, or “debiasing,” ¹ the next innovation is to harness the power of technology for what we call “pro-biasing.” To pro-bias, train algorithms selectively on controlled populations. For example, algorithms are now being used to make various judicial decisions in a number of US states, particularly to factor in recidivism risk. Although decision-makers assume that more information leads to better judgment, many of these risk models produce discriminatory decisions because calculations are based on historic crime data which has a track record of bias (Larson et al., 2016).

Corporations are already seeing the benefits of pro-biasing in algorithmic suggestions that drive sales while increasing consumer satisfaction. Consider YouTube, whose video recommendations capture audiences in a stream of increasingly extremist content, thus generating greater user engagement and advertising revenue (Tufekci, 2018). With careful planning, biases can be optimized, and algorithms can go from accidentally discriminatory to intentionally profitable.

An algorithm ensures that biases are difficult to detect, prove, or attribute to individuals: think of algorithmic blackboxing as supercharged red tape. Software is also easier to manage than having to foster and discipline cadres of loyal bootlickers. Willful ignorance does not apply because algorithms by default operate on a need-to-know basis. Even when biases are detected, they can easily be deflected by transferring blame onto incomplete training and advocating for more data. If datasets are demanded, the practice of data mining can be useful, that is, claiming the data is mine and thus, restricted. Algorithms allow leaders to feign ignorance of computer science and computer scientists to feign ignorance of subject matter expertise, automating their interests while obscuring responsibility, all in the name of neutrality and efficiency.

Build resilience through infrastructures

Achieving dominance at scale in the information age requires early investment in data infrastructures: the servers, storage facilities, transmission lines, offices, fuel, and labor that support the Internet. Building data infrastructure demands substantial up-front capital but yields massive value in the long term. Infrastructures are inherently resilient: generally unobtrusive and costly to dismantle, they fly under the radar and stick around. Appropriate infrastructures early to sustain appropriate activities into the future.

Ultimately, she controls data infrastructures controls information. When protests get too zealous, it is easy to shut off the internet on the grounds of protecting democracy (Access Now, 2022), but this heavy-handed strategy is both morally offensive and inefficient. A rights-based approach relies on the strategic production of better data. Rather than hiding or restricting information, resilient infrastructures ensure limitless production of quantitative data needing expert analysis, while stopping inconvenient data from existing in the first place. For example, increasing the density of pollution sensors in low-pollution locations, while lowering the density of sensors in high-pollution areas is an effective way to bend the arch of transparency to serve your needs (Xie, 2019). Whereas misreporting or deleting datasets can be easier to catch, the strategic placement of sensors can go unnoticed when raw data are accepted as objective.

Every data cloud has a silver lining. Big data is big business, with US$6.7 trillion locked up in the growing Internet Value Chain that ultimately relies on a growing telecommunications foundation (GSMA, 2022). Take advantage of tax breaks, slow regulators, and development incentives to get in early. Keep an eye out for chokepoints and opportunities for vertical and horizontal integration. Remember that data infrastructures also include people, and work diligently to reign your workforce. Learn from Amazon, whose proactive management keeps their labor force compliant and supply chains unstoppable (Streitfeld, 2021), while early investments enabled the company to clinch 33% of the cloud market share (Financial Mirror, 2022). Visionary leaders invest in the frontiers of technology while espousing the virtues of equality: tout the accessibility of blockchain, buy a bitcoin mine.

Make data publics work for you

The information age demands creative and dynamic engagement with digitally empowered stakeholders, or what is known as data publics. Whereas older modes of public relations sought to shut out the media, modern forms of messaging and vast datasets can promote active, but highly selective, engagement. Smart communications strengthen your base, undermine your opponents, and protect your interests from antagonistic members of the press.

Slick websites and social media optimization are important for mounting a strong rhetorical offense. Maximize gloss and minimize details. Don’t hold back from crowding out Twitter with keywords such as “transparency,” “good governance,” “open source,” and “hackathon.” Brandish diversity. Give a TED talk. In public-facing materials, prominently feature photogenic queer, black, indigenous, and other people of color. As an example of this can-do spirit, consider Facebook Oversight Board’s materials. ² Having only 3.9 % of the company’s employees identify as black has not deterred the company from featuring mostly people with multicolored skin tones on their promotional materials (Williams, 2021).

Whatever you do, call it “evidence-based.” For example, when faced with public pressure, agrochemical companies have competently linked their products to various evidence- or science-based sustainability standards (Freidberg, 2017). Across supermarket aisles, consumers encounter a bewildering array of sustainability stamps, green labels, and environmental claims that indicate compliance with environmental concerns while justifying a higher price. In this blur of standards, consumers glow in their consciousness while you can sell the same products for a premium. Rigorous analysis points to the value of more, not fewer, standards. Counterintuitively, an abundance of data and standards blunts trust in them all. Regardless of your business, we recommend adding several certifications to your product, complete with QR codes.

Although such strategies update analogue communications tactics, the real power lies in making data publics work for you. Provide controlled opportunities for your stakeholders to access and use data through a constant churn of experimental tools that bundle, synthesize, disaggregate, and visualize. Consider Chevron's chart generator, ³ a tool that provides an illustration of how interactive graphics can distract from negative coverage published elsewhere by making certain data sets available for public analysis. By allowing users to generate their own visually appealing accounts of corporate success from carefully curated data, such a platform updates public relations to the interactive age.

Conclusion

Above, we playfully invert key insights from STS scholarship on algorithms, infrastructure, and data publics to explore how those in power creatively exploit the particularities of data systems and the positive connotations around transparency to accumulate power and profit. By codifying operations for analyzing and managing vast quantities of information, algorithms reproduce, and amplify existing social biases even as they subject complex problems to the actuarial logics of a narrow coding elite (Benjamin, 2019; Burrell and Fourcade, 2021; Noble, 2018). The highly technical nature of algorithmic decision-making further complicates, rather than facilitates, the attribution of responsibilities for biases (Caplan et al., 2018). At the same time, data have a growing physical footprint, one apparent in burgeoning telecommunications networks and data centers, as well as the bodies that labor to maintain infrastructures that are expensive to build. Infrastructures are difficult to change once established and can elude critical inquiry given the rhetoric of data as immaterial and limitless (Pickren, 2018; Star, 1999). Finally, the satire shows how transparency technologies hail new sorts of subjects (Gabrys, 2016) or what Ruppert (2015) calls data publics. Hooked into the drama of public-facing transparency efforts—reflected in the constant release of new indicators, glossy reports, audits, meetings, and hackathons, this emergent data publics is busied with an exhausting process of analysis, rather than critique. By exploring how powerful actors use algorithms, infrastructures, and data publics to benefit and obfuscate, we challenge the assumption that data-driven tools are necessarily good for accountability.

Our unorthodox introduction of the concept invites empirical and theoretical investigations into how smart corruption works in practice. Although our satire adopts consultancy conventions to offer universal instructions to game accountability, as critical researchers we recognize that smart corruption will operate differently when carried out by distinct actors and in particular geographies. Empirical research should attend to these differences, examining who practices smart corruption, with what consequences, and at whose expense. Scholars may also consider how specific technologies differentially afford opportunities for gaming and how aspects of data systems—the black box quality of algorithms and the large capital investments required for infrastructures such as data centers, for example—enable specific techniques of gaming accountability. Finally, it is important to unpack the legal and political institutions that embolden gaming accountability as well as the contestations around smart corruption.

Calling out smart corruption can inform efforts to counter it. Because gaming accountability is difficult to identify or evaluate through existing ethical and moral standards, effective counteractions will require creative and critical thinking within relevant communities of practice. Deeper technical engagements with data-driven tools and the rhetoric of counting are already providing some solutions. Social organizations and movements are finding ways to use data to highlight and push back against structural inequality (Cinnamon, 2019; McElroy, 2020; Mrinali et al., 2022). This work too takes place in the arena of data publics, where visual and textual information is used to expose injustice and advance alternatives. At the same time, it is important to question the very construction of quantification as an obligatory passage point for advancing transparency and to consider what alternatives to data-centered accountability mechanisms look like. We offer this commentary as a modest contribution to collective efforts to imagine technology differently; articulating the gears of gaming accountability is another way of elucidating the levers for change.