Hacker News new | ask | show | jobs
by xg15 1009 days ago
> It has no legitimate interest in mandating statistical data-collection about online speech, either: because there's no legitimate government function that data serves and informs.

Not sure if this is an actual legal argument, but I don't see at all how the latter follows from the former - I think it would in fact be really bad if it did.

Statistical data collection isn't just necessary for the government to implement its current obligations, it's also necessary for policymakers (i.e. parliaments) to know which laws should be changed and which new government functions might be necessary.

I you don't like this for partisan reasons, it's easy to flip this around as well: Imagine some hypothetical federal government going all in on "no borders, no nations" and making it illegal for states to stop people at the US-Mexican border at all. By your argument, it would also automatically become illegal to even just monitor the border and count how many illegal immigrants are crossing.

Then later, when opposition and advocacy groups want to challange the federal government's policy, the government party can just say "the policy seems fine, we don't have any data that would show any sudden influx of illegal immigrants, so why should we change it?"

> the DeSantis government of Florida has mandated social media platforms disclose statistics about LGBT content on their platforms

There is an important distinction between data that makes someone personally identifiable and data that does not.

Not an expert in US domestic politics, but I believe that progressives don't generally have a problem with broad statistical data collection about marginalized groups - on the contrary, they often use this data for their own arguments, i.e. to show how unequal the distribution of men and women in leadership positions is.

What is a problem is any kind of data that would allow the government (or anyone else really) to individually identify persons from some group. (This includes "pseudonymous" data and very fine-grained statistics, because those can often be deanonymized)

So:

"How many % of SF's population are gay?" -> cool

"How many % of the tenants in Christopher Street are gay?" -> not cool

2 comments

- "Not sure if this is an actual legal argument, but I don't see at all how the latter follows from the former - I think it would in fact be really bad if it did."

There's balancing tests like the one in strict scrutiny [0] (IANAL/not actually sure if strict scrutiny applies here: it's just one example of a constitutional test relevant to the First Amendment). This weighs if a law that's potentially unconstitutional serves some important, real purpose that outweighs the tension it imposes on constitutional rights—a "compelling state interest" in the strict scrutiny test.

And there is tension with the First Amendment here. A law that merely "monitors" speech has a tangible chilling effect on the speech that's singled out for surveillance.

[0] https://en.wikipedia.org/wiki/Strict_scrutiny

The parent comment is using the word legitimate in terms of how the Constitution limits government power, not in the sense of the political beliefs of whoever won the last election.
even the constitution can be amended - or can have widely differing interpretations as the surpreme court is now helpfully demonstating.
Yes, but it's not a distinction without a difference.