Data cannot save us from ourselves

Clay Johnson, author of The Information Diet wrote on Twitter:

Redistricting should be done with data & open source software, not by humans.

Let me be completely honest: I don’t like golden calves and I can spot two in this single sentence: technology and data. Yes, redistricting should be done with data and technology, but the moment you add “not by humans” you ruin everything. The right sentence is:

Redistricting should be done by humans, using data and open source software.

Data-driven redistricting can prevent basic and shameless gerrymandering, but you can’t ask the computer to discover the “right” district structure because there is none.

If you want to algorithmitize everything you end up with a 1:1 map (remember Borges’ On Exactitude in Science?). And if you don’t, how do you select the “right” data? Meta-data? Meta-meta-data? Factor analysis? Cluster analysis? And what if the other people don’t agree with you? Don’t be naïve: you can’t remove the “human factor”, you’re just moving it around.

Data and open source software can add a layer of rationality and legitimacy to your decisions but you can’t ask the computer to decide for you. Be human, decide, and fight for what you believe in.

Pro tip: never use default formats.


2 thoughts on “Data cannot save us from ourselves”

  1. I could not agree more. Having the right data to answer the right questions, is not a task that can automate … yet. 😉 There is much informal information that can only be “decoded” by humans.

  2. I do not agree. Humans are mostly getting confused when a lot is happening in a short time. Data history can then help to create the right focus, by using pattern recognition.

Comments are closed.