Looking at Petrov’s (2020) ‘Impressive Big Data Statistics’, it quickly becomes obvious how much Data is generated daily. For example, 2.5 quintillion bytes (or 2.5 billion Gbs) of data is produced by internet users each day. While the sheer volume of data is overwhelming, it’s the practices to obtain, understand, use and distribute this data that that is the primary concern and main consideration of Data Ethics.
According to Accenture (n.d.), Ethics & ethics codes “set out the standard for acceptable behaviour within a profession”. It also identifies the standard of practice that society should demand from those within a particular profession. So, ‘Data Ethics’ outlines the standard, acceptable behaviours that data professionals should possess in all aspects of the ‘Data Supply Chain’ and data practice overall.
The mind-map below highlights the four key ethical theories (light blue) that can be linked to today’s data activities (dark blue), and how and where they could be applied. While each theory suggests principles that could influence data practice as a whole, I see some theories having greater relevance to some activities over others, and explore this further within each theory below.
Hypothetical Scenario to Apply Data Ethics
To explain each ethical theory plainly, I am proposing a hypothetical scenario and will suggest how each ethical theory could be applied.
Scenario: You are a Mark from the Marketing team at an Australian Bank, and you are collecting a variety of data/metrics from customers to develop tailored promotions that encourage them to spend more on their credit cards. The more customers that you can target, the richer the offers that your partners will allow you to promote, and therefore the greater opportunity for you to achieve your KPI’s of ‘driving revenue’ and ‘improving customer experience & engagement’.
The 4 Ethical Theories
1. Social Contract
The first theory is Social Contract, which suggests that “people live together in society in accordance with an agreement that establishes moral and political rules of behaviour.” (The Ethics Centre, 2016) It’s the formation of ‘social’ agreements, like laws or waiting in line at a checkout, that establishes the framework of how society functions.
How is this useful for Mark? — This theory sets out the respect you must-have for the customers, and the trust they have in you as a result. It ensures that you have internal practices/ processes that keep the data secure, private and anonymous, you are accurately analysing their data so that we are appropriately targeting the offers, and are transparent in the overall process to develop the campaign for the customers.
The issue with this theory is that the ‘social agreement’ of data is not as clearly defined as it is in society. Generally speaking, laws are fairly consistent across the world (theft, murder, fraud etc.), however laws of data practices are varied and inconsistent. Cookie Consent in Europe vs Australia is one example of this.
What’s not-so-useful for Mark? — You decide to use cookies to track customer’s movements and preferences when they log into their account online. You then use this data to optimise your campaigns, diversify your channels, and drive more spend engagement. In Australia, it’s not illegal to not disclose this or obtain consent from customers, so is it ethically wrong?
2. Virtue Ethics
The second theory is Virtue Ethics, which helps to define the ‘right’ actions that determine a virtuous person. It outlines the behaviours that are deemed good or bad and suggests that we should model our own actions on these behaviours to be a ‘good’ person (The Ethics Centre, 2016).
As Mark, from Marketing — having the customers consent to collect data, and communicating how these promotional offers are generated is important. Ensuring that you aren’t promoting to customers who are at high risk of not being able to pay their bills is an ethical practice and maintaining anonymity and privacy in the storage and access of this data is a major consideration.
The limitation of Virtue Ethics is highlighted when you are asked to decide whether to promote travel or retail-related offers. Travel spend may generate more revenue, but retail spend is the most popular amongst customers and will drive higher engagement. Neither decision is good or bad, in both scenarios the customers are still getting a great promotional offer, and you are still hitting both your KPI’s, so it’d be difficult to apply Virtue Ethics.
3. Kantian Ethics
Similarly, Kantian ethics looks at the intentions of an action. It proposes that “acts are to be judged by their accordance with universal principles rather than their good or bad consequences” (Oxford Reference, n.d.). It suggests that humans should never use another human to achieve a greater result. For example, it’s okay to eat food when hungry, but stealing is ethically wrong as it is taking from another person. (CFI, 2021)
What Mark thinks is good — This ethical theory ensures that you approach the campaign with good intentions from the start. You are respecting the customers and are ensuring that they understand your purpose in collecting their data. You are making sure that the customers know that you aren’t using them to achieve only your own goals; this campaign is benefiting both parties mutually.
The Corporate Finance Institute (CFI) also highlight out an interesting limitation of Kantian ethics,“Cheating on a test can only be moral when everyone else’s cheating on a test is justified.” So, cheating is morally okay if it is universally justified, but in reality it would lead to a failing of the education system’s integrity.
What does this mean for Mark? — you want to give the best experience to as many customers as possible, so you also target ‘high-risk’ customers, but promote fewer offers to them to limit their spend potential. You also know that other Australian Banks do the same with their campaigns and achieve high customer satisfaction as a result, so everyone does it- which makes it okay, right?.
Your intentions towards your customers are good, so this is ethical, however, 70% of the high-risk customers have not been able to pay their bills as a consequence, which doesn’t have an ethical impact under the Kantian theory.
4. Utilitarianism
The last ethical theory is Utilitarianism and its approach to ethics is different from the others. Rather than focusing on the intention or action, Utilitarianism looks at the effect of an action to determine whether it is right or wrong. (Nathanson, 2021). Tardie (2020) highlights that “Utilitarianism would say that an action is right if it results in the happiness of the greatest number of people”.
How does this apply to Mark? — you understand that if the campaign audience is larger, you will be able to send them better promotions, impartially benefiting both customers and the business. When looking at the customer base, you determine that if you alter the ‘risk weighting’ criteria within the data, within reason, you can increase your campaign audience by 30%, meaning your campaign will now reach some customers that have were deemed ‘higher’ risk of not being able to pay their bill. Now, you will be engaging more customers, with better offers and revenue potential for the business. You are achieving the ultimate ethical goal, resulting in the happiness of the greatest number of people, despite the risk of burdening more people financially. This highlights the benefit and harm of Utilitarianism.
Conclusion
Ethical implications and what is deemed as ‘ethical’ is complex, as highlighted by the benefits and shortfalls that each ethical theory has on our hypothetical Marketer ‘Mark’. Whether or not you are an ‘ethical’ data professional is somewhat dependent on which ethical theory lens you are taking. This is why ‘ethical’ data practice must be measured equally by all ethical theories when trying to evaluate whether a particular data practice is ‘ethical’ or not, and I see it as an essentual element in continued ethical data operation, innovation, and evolution.
References
- Petrov, C. (2020, September 10). 25+ Impressive Big Data Statistics for 2020. TechJury. https://techjury.net/blog/big-data-statistics/#gref
- Accenture. (n.d.). Building Digital Trust: The Role Of Data Ethics In The Digital Age. Accenture. https://www.accenture.com/au-en/insight-data-ethics
- The Ethics Centre. (2016, August 31). Ethics Explainer: Social Contract. The Ethics Centre. https://ethics.org.au/ethics-explainer-social-contract/
- The Ethics Centre. (2016, February 16). Ethics Explainer: Virtue Ethics. The Ethics Centre. https://ethics.org.au/ethics-explainer-virtue-ethics/
- Oxford Reference. (n.d.). Kantian Ethics. Oxford Reference. Retrieved January 25, 2021, from https://www.oxfordreference.com/view/10.1093/oi/authority.20110803100029911
- Corporate Finance Institute (n.d.). Kantian Ethics. Corporate Finance Institute (CFI). https://corporatefinanceinstitute.com/resources/knowledge/other/kantian-ethics/
- Nathanson, S. (n.d.). Act and Rule Utilitarianism. Internet Encyclopedia of Philosophy. Retrieved January 26, 2021, from https://iep.utm.edu/util-a-r/
- Tardi, C. (2020, November 23). Utilitarianism. Investopedia. https://www.investopedia.com/terms/u/utilitarianism.asp
