Which Measure? Whose Success?

Bill Gates believes he has found the solution to the world’s problems. Essential to “improving the human condition” is “feedback from precise measurement.” “You can achieve incredible progress,” according to Gates, “if you set a clear goal and find a measure that will drive progress toward that goal—in a feedback loop.”

Granted, as “mathbabe” Cathy O’Neil said in response, “given perfect data-collection procedures with relevant data, specific models do tend to improve, according to their chosen metrics of success.” But the metrics of success don’t choose themselves. After drawing out the politics that Gates obscured in several of his technocratic vignettes, O’Neil sagely warned:

[T]he person who defines the model defines success, and by obscuring this power behind a data collection process and incrementally improved model results, it seems somehow sanitized and objective when it’s not.

Don’t be fooled by the mathematical imprimatur: behind every model and every data set is a political process that chose that data and built that model and defined success for that model.

In 2010, David Rieff rebuked Bill Gates and similar-minded progressives for fetishizing “the idea of civil society as a kind of universal ideological solvent,” and believing that “in tandem with scientific innovation, the road to our collective salvation is now open to us.” On development, Rieff saw Gates as proposing solutions in “the purely technical sense.”

But development is not a software problem that can be resolved—as Bill Gates and Paul Allen developed new products for their corporation—by bringing the best minds together to brainstorm innovative [sic] solutions. Development is a matter of culture, of politics, and of justice, far more than it is a matter of technology or, for that matter, the technologized vision of human beings that can, without embarrassment, speak of ‘unlocking’ people’s potential as if they were seams of some precious mineral buried in the dirt.

In this vision of the world, “there are no great ideological contradictions, just issues of ’empowerment,’ ‘good governance,’ ‘transparency,’ and ‘accountability.'” But no amount of data or innovation can take the politics out of politics.

Before a “problem” can be recognized in politics, it must be defined. As the political scientist Steven Teles said, “it is through research, discussion, deliberation, and argument that we piece together the pieces of a complex world that we come to recognize as stable problems.” Worldly choices must be made about what data are relevant and how they should be modeled. In making policy, those choices are bound up with rival values and dueling interests. This reality is obscured by precious little stories where data can “drive” and spreadsheets can “speak,” as if they were Clippy.

update: Like a ventriloquist’s dummy, The Numbers never speak for themselves. As Kate Crawford explained:

Data and data sets are not objective; they are creations of human design. We give numbers their voice, draw inferences from them, and define their meaning through our interpretations. Hidden biases in both the collection and analysis stages present considerable risks, and are as important to the big-data equation as the numbers themselves.

Datasets may seem abstract, but, Crawford said, “they are intricately linked to physical place and human culture. And places, like people, have their own individual character and grain.” Raw data is an oxymoron.