A notable example is the way that gun violence data has been continuously and specifically suppressed by those who oppose gun control in the US (e.g: the Dickey Amendment, which had an extreme chilling effect on the CDC studying gun violence).

Unfortunately, it is common for those with power to see the data as a stick they can be beaten with rather than as a tool to do better.

Of course, an example of the negative potential also comes to mind: recently we got proof of extreme gerrymandering done using racial data. Obviously that isn't to suggest that data shouldn't exist, but more to underline your last point: fundamentally it can only be beneficial if people come to it in good faith.

>is the way that gun violence data has been continuously and specifically suppressed by those who oppose gun control

Last I checked they were happily singing out about the data because legal gun owners shoot statistically nobody and military style arms kill statistically nobody compared to career criminals with boring old handguns (who mostly kill each other because they can't rely on the courts to mediate their disputes).

I don’t see how this comment relates to my post or the broader conversation at all. This seems highly off-topic.

Well, my intent was to highlight that there is a similarity: those deciding how (or if) this data should be collected wish to suppress it because they believe it would result in change when they are invested in the status-quo.

Fundamentally trying to insulate data collection from those with invested with the status-quo (in all situations) would be sensible.

I'm not against collecting data. But one problem with mixing data collection and politics is that just by selecting which data to collect and which data to not collect you are implicitly crafting a narrative. It's very subjective, and the data collected will be based on the agenda of the politics at hand. There is no such thing as pure unbiased data collection.

It can be done though, can’t it? The SEC’s mandate is largely to enforce accurate and timely disclosure of information to the public. It’s not a perfect system but it works pretty well.

Why couldn’t there be an SEC-type entity that requires periodic public filings from police departments or hospitals?

This is certainly true in the abstract (we see a lot of this in homogeneous institutions like universities and newsrooms and it’s a real problem), but I don’t see the risk in this particular case. Further this is a generic argument against empiricism—it’s hard to get unbiased data, but that doesn’t invalidate the enterprise.

> Our national debate would be so much better if we had better data.

I doubt it. People have crunched the data that we do have and found that it doesn't support the popular narrative - and been fired from their jobs just for crunching the numbers.

If you could cite a source or even what you think the popular narrative to be that would be great.