Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Noise infusion banned from statistical products published by Census Bureau

desfontain.es

Post reply on HN
Print view View original

531–540 of 662 posts

Re: Noise infusion banned from statistical products published by Census Bureau

#531

Post by Forgeties79 »

appreciatorBus wrote: post #485
Forgeties79 wrote: post #328

Earlier quoted context omitted.

> If data about the public is so dangerous that we must disguise the results, then perhaps its data we shouldn’t be collecting in the first place. By this logic no one should ever collect your address for any reason ever. How do we function as a society if we can’t ever give PII in any context? Anonymization/security is critical and makes a lot of critical functions possible. How could you receive your mail in a worl…

This is silly. An argument about whether or not to deploy differential privacy on large statistical databases has no bearing whatsoever on whether or not you give your address to have a package delivered. If you want the package delivered, you have to give your address. On the other hand, it’s not at all clear that people should have to involuntarily, my force of law, offer up all sorts of personal details about thei…

It’s not silly. I’m responding to this:

> If data about the public is so dangerous that we must disguise the results, then perhaps its data we shouldn’t be collecting in the first place.

We agree that doxxing is dangerous online yes? Your point about the white pages is exactly what I’m talking about. A piece of data isn’t inherently dangerous or not dangerous. It’s about context and ease of access by actors with various intentions.

Re: Noise infusion banned from statistical products published by Census Bureau

#532

Post by nl »

foolfoolz wrote: post #68

i have such a hard time reconciling stuff like this: > The census bureau decided to adopt differential privacy for the 2020 Census and: > The consequences will be dire for utility or for privacy, and possibly both. It's hard to understate this point: future statistical releases will either be useless compared to past ones, or they will be incredibly unsafe so we took the census for centuries before this point, and it…

> so we took the census for centuries before this point, and it was “ok.”

It wasn't ok - it's been shown that the data released could individually identify people in releases before the 2010 Census.

Re: Noise infusion banned from statistical products published by Census Bureau

#533

Post by drysine »

brainwad wrote: post #252
drysine wrote: post #242

Earlier quoted context omitted.

[flagged]

As I understand it, in the Soviet Union you had to get government permission before moving. Whereas here the right to move is guaranteed, you just have to update your details.

You are right, I apologize

What's the fine for not registering your move? What time do you have to update your location?

In Russia in the 90s we had 5 days and now it's three months. The fine is 3000 rubles (~40 euro) and zero if you moved in the same region (most of them are bigger than Switzerland:) or live at relative's or spouse's place

I used to think that what we have in Russia is Soviet legacy (albeit relaxed) and it's something people in Europe are not burdened with

Re: Noise infusion banned from statistical products published by Census Bureau

#534

Post by nandomrumber »

jibal wrote: post #519
nandomrumber wrote: post #508

Earlier quoted context omitted.

You should never trust a politician, especially one that claims to be on/in the left. These two statements are equally as vacuous.

No, the first one is true. It's the copycat comment that is vacuous and doesn't even make any sense -- "in the right" has a clear meaning, "in the left" is nonsense.

[dead]

Re: Noise infusion banned from statistical products published by Census Bureau

#535

Post by tgv »

kccqzy wrote: post #462

An earlier article on the same blog had some very useful information on how easily the aggregated census data from 2010 (before differential privacy) could be used to reconstruct real data for individuals: https://desfontain.es/blog/us-census-reconstruction-attack.h... I am personally convinced that the reason noise infusion was banned was because powerful people were already reconstructing individual data from censu…

Or worse. It allows anyone to build really targeted data sets. Insurance companies would love such data, and many of those will use them without scruples.

Re: Noise infusion banned from statistical products published by Census Bureau

#536

Post by Forgeties79 »

pessimizer wrote: post #311
pelorat wrote: post #129

Earlier quoted context omitted.

Sure, in Europe we don't because we already have databases of all citizens, also recording attributes like race, skin color, religious affiliation or political leaning in a database is highly illegal, both for the government and for private use.

> recording attributes like race, skin color, The only reason we ever started doing this was to track ex-slaves and their descendants, and after-1965 every other possible grouping of people started begging for a category that it could use to get government grants in some way. The irony is that now, when censuses somehow desperately need to figure out if you're Armenian or not, they don't count the descendants of slav…

> The only reason we ever started doing this was to track ex-slaves and their descendants, and after-1965 every other possible grouping of people started begging for a category that it could use to get government grants in some way.

Both of these comments need citations. The first I can maybe buy but the second is harder to accept without proof.

Re: Noise infusion banned from statistical products published by Census Bureau

#537

Post by Zoombini »

Dylan16807 wrote: post #323
Bratmon wrote: post #92

Earlier quoted context omitted.

But this article is about a decision to damage the census less. If you value an accurate census, you should be celebrating!

If they follow the rules, preserving privacy via cruder methods, the data will be much more damaged. For any particular level of privacy, the banned methods can give you more accurate data. For any particular level of accuracy, the banned methods can give you better privacy. The only way we're getting more accurate data is if the new rule causes them to largely abandon privacy. That would be bad. Harm for no benefit.

Data collected by public entities or with funding from the public should be expected to be published in formats digestible by the public. There should be no gatekeeping of the source data.

Re: Noise infusion banned from statistical products published by Census Bureau

#538

Post by DaSHacka »

fc417fc802 wrote: post #526
Loughla wrote: post #492

Earlier quoted context omitted.

That's a very strong statement. I'm absolutely no fan of ICE or indefinite detention of people based on where they were born, but I'm not sure we have camps where we're committing genocide in the states. Nor do I see that in our future.

So if the death camps hadn't been for the purpose of genocide but rather for disposing of the poor or convicted criminals or perhaps those suspected of being illegal immigrants would that have made them okay? It seems to me that "I'm not sure we have camps where we're committing genocide in the states" is largely irrelevant to the issue at hand.

You're making the argument the Nazis only deported people?

Sounds ironically more generous than the actual nazis claiming it either didn't happen or was 'deserved'.

Re: Noise infusion banned from statistical products published by Census Bureau

#539

Post by DaSHacka »

Alive-in-2025 wrote: post #305
dnautics wrote: post #291

Earlier quoted context omitted.

There is a significant movement in conservative circles that "the census should literally only be a count". this could be a wedge to prevent detailed demographic data collection by the government

If you don’t count people of different races, nationalities, origin, then you can pretend there are only white people. You don’t know if there are any marginalized people gathered in say an extremely poor neighborhood, because they are all people. I see it as a way to pretend there’s just white people.

Why would a conservative want to pretend that? When was the last time you actually talked to one?

Re: Noise infusion banned from statistical products published by Census Bureau

#540

Post by well_ackshually »

whimsicalism wrote: post #500
well_ackshually wrote: post #143

Earlier quoted context omitted.

France used to make plenty of lists. We loved lists. Lists are good. Jews lists? Sure, it's maybe useful one day when we want to do something. Boy were the Germans happy to find these. The American obsession with asking for people their perceived origins (AAPI, AA, Latino, ...) is more than weird: it's downright dangerous. Don't fucking ask these questions, and never, ever write it down, especially not with names. Th…

France knows very little about managing a post-colonial multiracial society (except for terribly), I would appreciate if y'all listened and learned or at least approached the issue with more humility. France has serious racial, colonial, Muslim, immigrant, and banlieue inequalities, but its refusal to officially measure race/ethnicity makes those inequalities harder to see, litigate, quantify, or remedy.

Considering the last time we made lists of people's origins, it was used by the nazis and that there's a chance for the neonazis in 2027: no, thank you.

Knowing someone is algerian, muslim or black doesn't help you fix inequalities. It doesn't help in the US, or anywhere else in the world. Racial statistics are useless. We know where poverty is.

Post reply on HN