Comment by 1propionyl

The catch-22 is that being data-driven means being driven by what data you can measure. And that is an absolutely gobsmackingly fatal flaw.

For a 1000+ page exposition on why this is such a fatal flaw see James C. Scott's 1990s tome "Seeing Like a State" which addresses precisely this issue with receipts brought from fields ranging from agriculture, urban planning, politics, family naming systems, and more across at least 4 or 5 continents and a century.

It's an essential non-partisan read that manages to piss everyone off by being a peon to left anarchism in many ways while saying "actually yes Hayek was right about sensitivity to local conditions" and "actually the NHS and NWS are good and sometimes centralized planning isn't the worst solution".

I've not seen any direct link between them but I feel like you can draw a very firm line between it and Piketty's Capital and Ideology. You might think "wait isn't Piketty the guy who's completely data driven and publishes stupidly large amounts of xlsx files as appendices" and you'd be right except in his second book he's leaning far more heavily on sociology and just using financial data (much imputed from tax records) as support for sociological arguments that align very closely in many respects.