←back to thread

Experts warn about the 'crumbling infrastructure' of federal government data

(www.npr.org)
99 points xnx | 2 comments | 15 Mar 25 03:41 UTC | HN request time: 0s | source
Show context
KennyBlanken ◴[15 Mar 25 04:46 UTC] No.43370057[source]▶
Destroying agencies that collect statistics is a "feature", not a bug.

Why deal with unemployment, declining GDP, etc when you can just shut down the agency figuring out how many people are unemployed and where the GDP is going?

Better yet, outsource it to the guy who gave you $10M in campaign donations or bought your crypto coins and then wink-nodded at you, and give him a $50M contract to do something the feds were doing for $20M - and he lies to make you look better. Win-win-win...

replies(5): >>43370097 #>>43370177 #>>43370303 #>>43370433 #>>43373483 #
1. Centigonal ◴[15 Mar 25 06:21 UTC] No.43370433[source]▶
The article covers many cases that predate the current administration. I'm not contesting your point, but I think TFA is about a bipartisan failure to keep federal data collection infrastructure up to date.
replies(1): >>43371834 #
2. uppost ◴[15 Mar 25 11:31 UTC] No.43371834[source]▶
It's not a very good article, but it hints at another angle, the problem is not only lack of funding growth, it is that modernizing and keeping up with the ever expanding mandates is not possible without major reinvestment. If they want great precision about Guatemalen migrant single mothers in the CPS because that is now a important group to gather data on, it costs a lot. (Worth noting too that these expanding mandates are often very political in nature and actually undermine the enterprise as adding too many features to software without regard to tech debt will often tend to do.)

What should strike many readers: zero mention of efficiency gains or emerging methods made possible by LLM, uav, and other modern tech. A tech savvy reader should wonder about that stuff.

One of the best sales pitches for increasing funding is specifically this, modernizing costs money in the next 5 years but saves much, much more money longer term. Is it not the right time for robot census workers?