Thinking About Ethics and Data

Tue, 20-Dec-2022

👨🏻‍🏫 A key dimension in our relationship with data is the one that doesn’t have a law behind it. What are our ethical obligations to ourselves, our team, our employer, our profession, or the public? Who decides? How do those obligations change based on the role we play?

What does it mean to live by a code of ethics?

Ethics are expected practices and behavior that define boundaries for our work and provide a framework for making “the right” decision. Ethics are typically standardized in a code of ethics, code of conduct, or standards of professional practice, and are enforced by the group rather than being enshrined in law. Examples of professional organizations that police their members include the American Medical Association (doctors), the American Bar Association (lawyers), and the Association for Computing Machinery (computer scientists and engineers). 

These organizations require that members adhere to established codes of conduct or risk expulsion. Members of a profession may practice without membership in a professional organization but run the risk of diminished standing within the profession. Some positions may even require membership in a professional organization as a condition of employment, which means real risk to the individual for failure to follow some standards of practice.

What organizations or standards of professional conduct exist for data engineers and analysts? That’s the question that we explore here.

A Code of Conduct for Data?

The State of California has many laws that protect individual privacy and laws that regulate what is permitted by State employees or elected officials. One example is the Conflict of Interest (COI) declaration required of designated employees and elected officials in financial or decision-making positions (a.k.a. “designated” positions).

Designated personnel must declare the source of their income and assets annually on a Form 700 and are required to recuse themselves from any contracts or decision-making where they have a financial stake in the outcome. The Form 700 enshrines an ethical obligation explicitly into law. Many ethical standards, however, are not backed by laws.

Where might we look for ethical guidance as it relates to our interaction with data? We need to explore further.

Do No Harm

The Hippocratic Oath is one of the oldest examples of an ethical code. The imperative to “Do No Harm” requires doctors to act in the best interest of their client and make decisions regarding their care that advance the interest of the patient over personal preferences, religion, or political affiliation. Without this oath, would doctors be more inclined to take shortcuts that benefit their wallet at the expense of your health? Would you trust a doctor that didn’t adhere to the Hippocratic Oath?

People that work with data  also have a similar obligation to “do no harm”. There is a level of trust that is given to data practitioners to tell the truth, objectively. System builders have access to build systems that advance the interest of their clients, but should they do so if it results in negative consequences that are born by someone else? How does one decide which interests have priority?

Even without an ethical code of our own, might we start with “Do No Harm”? Next Step: follow the links below.              

Read More

Step One: Laws — Make sure you know the State and Federal laws related to data handling. If you live in California, you may find the information provide by the State of California’s Health & Human Services Agency to be extremely detailed and informative:

Step Two: Ethics — Take a look at these existing codes, from the Data Science Association and the Association of Computing Machinery (ACM):


About the Author