I didn’t know rewards were common but have never looked into that, who reports those numbers even? They sound fun to go through
I didn’t know rewards were common but have never looked into that, who reports those numbers even? They sound fun to go through
dependencyinjection@discuss.tchncs.de 19 hours ago
I asked an LLM although many here hate that.
It basically said it’s very fragmented but many departments / organisations do publish their rewards information.
SEC, DOJ, Dept of state RFJ, FBI, DEA, Crime stoppers, and more.
You have to go and look at each one and consume that data ideally programmatically. The LLM has links you can try but I won’t post the output without consent.
bdonvr@thelemmy.club 15 hours ago
Why bother asking the bullshit machine
dependencyinjection@discuss.tchncs.de 10 hours ago
Why bother questioning what others do. Move a king you silly cunt.
NotForYourStereo@lemmy.world 16 hours ago
Imagine not being able to do your own research.
dependencyinjection@discuss.tchncs.de 10 hours ago
Imagine. I’m not a fucking researcher. I doubt you could do your own research.
NotForYourStereo@lemmy.world 9 hours ago
It’s really not that hard, buddy.
At least, not for those of us who aren’t absolute idiots.
LifeInMultipleChoice@lemmy.dbzer0.com 19 hours ago
I want all of the data. I see nothing wrong with LLM’s scraping government sites
dependencyinjection@discuss.tchncs.de 18 hours ago
I’ll paste the output. My prompt is below too so if you want to probe it for the best way. You could get it to make a program just have to keep pushing if stuff doesn’t work. I’m a developer so it helps if you know a little.
Prompt.
Where can I find data on reward money offered vs paid in the USA over the last 20 years
Reply Short answer — there’s no single complete national dataset.
Reward programmes are fragmented across federal agencies, state/local Crime Stoppers chapters, private bail/bounty firms, and corporate/financial whistleblower schemes. To build a 20-year “offered vs paid” series you’ll need to pull from several sources and (optionally) stitch them together.
Below are the best places to look (with examples and why they matter): 1. Federal whistleblower & reward programs (high-quality, machine-readable annual reports) • SEC — Office of the Whistleblower: publishes annual reports with amounts awarded each fiscal year (detailed PDFs). Great for securities whistleblower payments (large, consistent dataset since 2011).  • DOJ / Department of Justice press releases and program pages — DOJ has various reward/whistleblower programs (and a recent pilot program). Useful for criminal/corporate enforcement reward totals where reported.  • U.S. Department of State — Rewards for Justice (RFJ): posts individual reward offers and announcements (large, occasional awards). Good for transnational/terrorism-related reward offers.  2. Federal law enforcement (FBI, DEA, etc.) — press releases and “rewards” pages • The FBI and some other agencies publish individual reward offers (Ten Most Wanted, fugitive rewards) and occasional summaries. These are event-level (not neat annual totals), so you’d scrape press releases or file FOIA requests for bulk data.  3. Crime Stoppers (national + local chapters) — aggregated rewards paid • Crime Stoppers USA and many local chapters publish aggregate “Rewards Paid” and “Property Recovered” statistics (useful for local/community reward payouts). There are many chapter reports you can scrape/collect.  4. Private-sector and industry sources (bounty/bail industry, private reward platforms) • Bounty hunters / bail bond industry earnings and private reward platforms are private and fragmented — statistics exist in trade pieces, law reviews, and industry blogs but no central public dataset. See law reviews and industry reports for estimates.  5. News archives & academic evaluations • For historical context and one-off large reward offers (and sometimes totals), use major news archives, NCJRS/NIJ and academic reports (e.g., evaluations of Crime Stoppers). 
Practical approach to build the 20-year series
If your goal is an actual dataset (offered vs paid per year), I recommend this pipeline: 1. Decide scope — which kinds of rewards? (e.g., federal law-enforcement offers, whistleblower awards, Crime Stoppers/local rewards, private bounty/bail payments). 2. Pull high-quality federal reports first — SEC whistleblower annual reports, DOJ/State press release archives, FBI press releases. (These are the easiest to quantify.)  3. Scrape Crime Stoppers chapters — collect annual “rewards paid” numbers from national and major local chapters (they often publish totals).  4. Supplement with FOIA or API calls — for FBI/DOJ bulk records or if you need all reward offers (not just press releases). 5. Handle private/industry data as estimates — compile from industry reports, law journals, and news; treat these as less reliable.  6. Normalize & document assumptions — e.g., currency (nominal vs inflation-adjusted), whether “offered” is the maximum stated offer or the amount actually committed, and whether “paid” includes partial payments.
Useful starting links (to grab data now) • SEC Whistleblower Annual Report (FY2024 example).  • SEC newsroom (whistleblower award press releases).  • State Dept — Rewards for Justice program pages.  • FBI reward announcements / press releases.  • Crime Stoppers USA / sample local chapter stats pages. 
⸻
If you want, I can start pulling and assembling a first-pass dataset for you right now (for example: SEC whistleblower totals per year + Crime Stoppers national totals + notable federal reward offers per year) and deliver a CSV with year, source, offered (if available) and paid. Tell me whether to include or exclude (a) private bail/bounty payments, and (b) whether to inflation-adjust amounts — I’ll assume nominal USD if you don’t specify.
Goldmage263@sh.itjust.works 17 hours ago
Wow, yeah. People hate that you used an LLM, lol. Look at those down arrows.
rc__buggy@sh.itjust.works 16 hours ago
When it’s actually a perfect use of an LLM.
Some motherfuckers just insist on using hammers to drive screws
AwesomeLowlander@sh.itjust.works 15 hours ago
Right? This is one of the perfect examples where we don’t care too much about the accuracy of the data, it would be extremely time consuming, etc etc.
dependencyinjection@discuss.tchncs.de 10 hours ago
Yup. We are the left wing The Donald. They hate everything. And talk about the same shit incessantly without actually making any change cause they won’t engage with anyone else with a different opinion.