An American Privacy Emergency

Join the discussion

• Hacker News
• There has been debate among statisticians and political scientists about using differential privacy for census data. 2020 was actually the first Decennial Census that used differential privacy. This is the mandated census done every 10 years that counts population and is used for apportionment. Some have criticized the use of differential privacy.[1][2] But others have argued that coarsening does not protect privacy sufficiently and that differential privacy does not distort apportionment.

  The political context is unclear. There are lawsuits about whether differential privacy is constitutional. There is also the possibility that citizenship status can be inferred by using multiple census products put together. It's also possible redistricting is at stake although it's unclear to me how getting rid of differential privacy benefits any one party.

  [1]: https://apnews.com/article/business-census-2020-technology-e...

  [2]: https://www.science.org/doi/10.1126/sciadv.abk3283

  by onetimeusename
• I'd like to emphasize that coarsening is not just theoretically non-private, a number of attacks that lead to leaking personally identifiable data were demonstrated on the 2010 census. So it's not really a he-said/she-said situation.
  by _alternator_
• Yeah, calling your legislators is going to do precisely nothing [1], just like data centers are almost universally opposed by the communities and the negative externalities are way more real and direct. Yet they keep getting approved anyway.

  The true crisis here is in the captured political system.

  In the 1990s in Australia a racist, white supremacist party arose called One Nation through a very weird confluence of events that led a racist fish and chip shop owner by the name of Pauline Hanson to become a member of parliament. It was almost 30 years ago she gave her now famous miaden speech to Parliament [2].

  After some scandals, One Nation kind of disappeared for awhile, in part because the conservative coalition (of the Liberals and Nationals) basically adopted the racist platform in the early 2000s where asylum seekers were effectively scapegoated. But weirdly she's back now. Anyway, that part isn't the point.

  Australia has a preferential voting system, what tends to be called ranked choice voting in the US. You generally have two options on how to vote: you can individually number candidates yourself or you can use the registered preferences for a given party. In this case you put a "1" in Australian Labor Party, Australian Greens or whatever. A lot of people do this so preferences matter. Anyway, One Nation had a strategy of voting gainst the incumbent with preferences. So if it was a Liberal seat, the preference went to Labor and vice versa. This scared the bejsus out of the political establishment such that the opposing political parties gave preferences to each other over One Nation, leading to One Nation getting no seats in Parliament despite getting 10%+ (at its original peak) of the popular vote.

  My point here is that too many politicians and political parties view their seat as something that belongs to them. In the US primaries are treated largely as a formality by the parties for their anointed candidates. Re-election rates in Congress have sat at 95%+ for decades.

  What's interesting is that the Demoratic Party is almost in open revolt currently and over the past few weeks, several long-term (10-30 years) incumbents have been primaried by insurgent candidates.

  Here's a funf act I learned this week. It's been ~18 years since Citizens United basically got rid of campaign spending limits. A third of all the money spent since then has been spent this year on primaries. Thomas Massie has $35M+ spent against him in his primary, making it the most expensive in US history. Many others are in the millions. It's estimated that the total spending for the Senate seat in Maine will push $400M. For one Senate seat.

  All of this is a long way of saying that the only thing that will work is making these legislators fear they'll lose their cushy positions. And really if somebody has sat in office for 30 years and has nothing really to show for it, it's time for them to go.

  [1]: https://act.represent.us/sign/problempoll-fba

  [2]: https://www.youtube.com/watch?v=p2ypTX9ntTQ

  by jmyeet
• ask tom s. in california if all that spend guarantees anything.

  more urgent is to repair the broken election process especially in california where it now takes 30+ days to "count" the votes.

  by onetokeoverthe
• Agreed: sure, call your representative. If they're cagey or noncommittal, do what you can to get their ass primaried. Every "moderate" will absolutely sell you into a Panopticon.
  by idle_zealot
• Ranked choice voting would go a long way, the two-party system is an intentionally forced false dichotomy like when parents give their kids the choice to eat broccoli or carrots so they’ll think it was their decision. Both parties are controlled by the investor class.
  by an0malous
• Heya, fellow Aussie here. Have you ever tried contacting your local MP?

  I was cynical at first like you, thinking why bother. But when I tried it, turned out I was wrong and I actually had a pretty good experience!

  The way I see it now, is that MPs aren't always in a good position to get close to the facts, so when you get in touch and tell them what you think.. you're actually giving them a huge gift.

  It can actually be pretty effective, especially for state/local issues. For federal stuff, sure, might not be as good, but you'll at least get some satisfaction from getting an acknowledgement from their chief of staff or secretary.

  by dhotson
• One of my favorite recently learned facts about Congress:

  Federally mandated parental leave (paternity and maternity leave) polls at about 80% in favor with the US adult population. This is regardless of political affiliation, by the way. Democrat and Republican voters both support it.

  Upon reading this, you might be surprised as to why it's NOT federally mandated given how popular it is.

  One group it's NOT popular with is corporations. And corporations donate a lot of money to politicians. And it's cheaper to donate to politicians who are against parental leave than it is to pay people for that parental leave.

  I enjoy sharing this b/c it's a reminder that there are groups who spend a lot of time and money to get their way. At first, that might feel overwhelming. You might be surprised to know that when you call your local congressperson, those calls gets tallied b/c they want to know what their constituents care about. So give them a call and let them know.

  by alexpotato
• if DSA gets enough people in, that might go through?

  pretty directly within the realms of what their candidates support, and they have a pretty good purity test to tell who to support or not with the genocide question

  by 8note
• If a million people want it and it is worth $100 to them they could create a superpac for this with $100m funding?
  by hahahaa
• Free housing, free food, free health care, and free income are also wildly popular with the US adult population. The problem is that those things are not really "free" because somebody else needs to pay for them.

  https://en.wikipedia.org/wiki/Tyranny_of_the_majority

  by anonymousiam
• Well, if you think money in politics or corporations buying politicians is bad now, it is going to get exponentially worse in the USA. The Supreme Court recently gave a decision that allows the rich oligarchy to give unlimited amounts of money[1] to their favorite puppets... excuse me, politicians.

  [1]: https://www.npr.org/2026/06/30/nx-s1-5827039/supreme-court-c...

  by baranul
• > Federally mandated parental leave (paternity and maternity leave) polls at about 80% in favor with the US adult population. This is regardless of political affiliation, by the way. Democrat and Republican voters both support it.

  This sounds like the sort of thing where if you poll people yes/no, a majority of people support it, but then if you ask them whether they'll change their vote over it, the people who said yes say it's not their most important issue and they don't care about it enough to do anything. Whereas the people who said no are e.g. small business owners if you don't exempt them from it because they can't afford to pay someone for extended time off while simultaneously paying someone else to fill in for them, or large business owners if you do exempt the small businesses because they hate anything that gives smaller companies a relative advantage, and both of them will actually care if you try to do it and oppose the policy either way for fear that it might get changed during or after being enacted.

  So then you have 80% wants x 1% cares vs. 20% wants x 75% cares and the second number ends up being bigger.

  by AnthonyMouse
• If we had voter referendums at the federal level, most “hot button” partisan issues would be solved because there is consensus across 70-75% of the population, even if weighted by state somehow
  by yieldcrv
• From the responses I've gotten from my representatives when I've written them, my impression is they care a lot more about their corporate sponsors and the party line than they do about their constituents.
  by thayne
• One of the things I have found so alarming about a lot of recent revelations is just how cheap congress goes for.
  by Fomite
• This has been studied, famously by Princeton [1]. The chance of any given bill passing is roughly 30% and any amount of public support from 0% to 100% has almost zero impact on that 30%.

  Elected offices have become fiefdoms to enrich oneself and maintain the status quo. Anyone who bucks this trend has historically rarely gotten into office or been chased out once they do. This could be from funding another candidate, simply starving an existing candidate of campaign funds or in some cases by redistricting somebody out of a seat.

  And look at the reelection rates for Congress [2]. They tend to hover between 90% and 95%.

  [1]: https://act.represent.us/sign/problempoll-fba

  [2]: https://www.opensecrets.org/elections-overview/reelection-ra...

  by jmyeet
• Related: https://news.ycombinator.com/item?id=48517377

  It's too bad this has become political.

  I do differential privacy work for GDPR compliance and it's an interesting technology.

  by nl
• > I do differential privacy work for GDPR compliance and it's an interesting technology.

  You mean legislation?

  by giancarlostoro
• Previous discussion: https://news.ycombinator.com/item?id=48517377
  by greyface-
• The article presents two possible methods to protect privacy in a dataset. It then attacks a theoretical weakness in a contrived scenario in the old method, which intends to incline us to choose the other, newer solution. The article does not describe in detail the newer solution beyond its name. I have have questions that the article doesn't cover in enough detail: 1) Has the coarsening failed in the way described in the article in practice which leaked information? 2) How does the 'other' solution we're expected to desire work? 3) What is an example of the difference in level of detail offered by the newer solution that was not possible when the data needed to be coarsened in practice?
  by jhhh
• (1) "A Simulated Reconstruction and Reidentification Attack on the 2010 U.S. Census" https://arxiv.org/pdf/2312.11283

  (2) By adding carefully tuned Gaussian noise. In the last 6 years we have also figured out how to add much less Gaussian noise: "The 2020 Census Disclosure Avoidance System TopDown Algorithm" https://arxiv.org/abs/2204.08986

  (3) This one is harder to answer, since the Census Bureau aimed to release the same style of statistics as in previous decades. So the goal of 2020 was to release the same statistics with the same error bounds. Evidence suggests they succeeded in doing this. "Evaluating Bias and Noise Induced by the U.S. Census Bureau's Privacy Protection Methods" http://arxiv.org/abs/2306.07521, "Evaluating the Impacts of Swapping on the US Decennial Census" http://arxiv.org/abs/2502.01320

  by flowercalled
• Can anyone explain me why the Heritage foundation targeted these statistical techniques? What's the political motive behind it?
  by stymaar
• empower big corporations, and blow open statistical protections to allow for 1) better agit-prop efforts, to allow for more easy chasing of minorities they don't like, and 3) to target elections
  by red-iron-pine
• Here's an example: https://x.com/WadeMiller/status/1985183761957372286
  by censusnerd
• The 2020 Census had a number of issues and many of them ended up giving the Democrats more seats in the House than they should have. Oh, and the Census results were due at the White House in December 2016, but somehow they didn't reach the WH until January 21, 2017.

  Was Differential Privacy specifically involved? I don't think so, but I think it's being lumped into the mix and being blamed for the other issues.

  by xhkkffbf
• Probably related to things like "The Nazis utilized data from routine censuses, tax returns, and municipal police registrations. In Germany, and in occupied countries like the Netherlands, this information was systematically organized. In some instances, IBM technology (via Dehomag punch card machines) was used to tabulate and sort census data to identify individuals of Jewish descent."
  by cma
• > If followed, this order will destroy the Commerce public data our nation relies on for important decisions, such as where to build necessary services for our community’s well-being

  So this is not about privacy. Scott sounds like a computer scientist forced (by the American ecosystem) to become a bombastic talker.

  by zkmon
• This is guest post by Cynthia Dwork; that's Dwork's writing, not Aaronson's.
  by Sniffnoy
• Can’t release data sets if they can’t be sufficiently anonymized so as to avoid Privacy Act/other non-disclosure statutes. Can’t anonymize data sets sufficiently if you’re banned from using the techniques data people use when anonymizing data sets. Not that difficult to follow.
  by dmvdoug