Thread by @whstancil

I guess ultimately one thing that bugs me about the current political discourse goes beyond specific ideas and people and is really about something more fundamental, which is the idea that it makes sense to rely very heavily on data and statistics as prediction tools in politics

It seems to me that reliance on data, especially left-of-center, has produced a lot of artificial certainty about "correct" political decisions, and created the impression we can predict effects of decisions with precision - but in reality, such predictions often fail completely

It makes sense that data tools come up short! You're talking about complex systems of tens of millions of actors all affecting each other, often in ways that are ill understood. Like predicting the stock market, but vastly more potential outputs. The tools aren't up to snuff

Even when we have data about WHAT ALREADY HAPPENED, we usually are unable to say with certainty WHY those things happened.

Then you shift the timeframe forward to the future, to a society that doesn't exist yet, and you want to say both WHAT and WHY, and it breaks down totally.

It ought to be humbling to think about how consistently terrible we are predicting things, and how deeply insufficient our data and statistical tools are for the task. An important part of rationally approaching a problem with tools is acknowledging their limitations and failures

Needless to say, though, that's not really what's happened. Sadly, the political world is NOT full of statisticians listing the near-infinite number of things they do not know, and the future events they cannot accurately predict.

What we've gotten instead is a kind of data theater, where various idea already in circulation are dressed up with numbers. It gives them more credibility among people who don't quite understand the numbers, who assume they're reading science instead of guided guesswork.

That's really damaging! It's created a high priesthood in politics, purported keepers of secret data knowledge, with disproportionate influence. But the priesthood's views often seem less grounded in empirics than in the reality that its members usually share a similar background

Meanwhile, in the real world, human beings still seem pretty capable of out-campaigning Deep Blue.

Intuition, creativity, common sense, and just wild experimentation seem as effective as - and sometimes better than - the data priesthood's purported empiricism.

The problem is that the social allure of smartness - again, especially left-of-center - has let the data priests monopolize a lot of political thinking, shutting out people with different, more honest, more flexible assessments about what to do.

That immiserates our politics, narrows our conversations, and pushes to the margins a lot of people who don't have the right background or don't talk a certain way -- all while the benefits for everyone but the data elite remain awfully difficult to discern.