🧵

I think it's very interesting that Erlang also saw a drop in salaries while F# and Haskell both saw a climb. There's a lot of variability in a number of languages. Why have Julia salaries dropped, for example? Most mainstream languages saw a 5-10% rise. Niche languages mostly much less. I suspect Clojure has seen the loss of several small, highly-paid teams -- based on what I've seen in #C050UG324GM and elsewhere -- and it's worth noting that the overall percentage of developers using these niche languages is very small (1.26% Clojure, 0.99% Erlang, 1.15% Julia, 0.97% F#, even Haskell is only at 2.09%) so it doesn't take much of a change in the composition of the developers taking the survey to have an outsized impact on some of the numbers. But, yes, it's also possible that more aggressive hiring of junior developers into Clojure shops has had an impact and it's also possible that geography has an impact: US salaries overall are very high compared to several other parts of the world -- an increase of Clojure developers in LatAm or Asia is also likely to bring average salaries down in niche market. SO's survey shows India (backend dev $21k vs $165k in USA) but doesn't show LatAm, China, or any Eastern European salaries which are all going to be lower than USA.

(I should have added Zig to the market share stats: just 0.83% -- 11th from bottom in usage so the salary variance is pretty much margin-of-error stuff)

Anecdotally, I’ve seen some Clojure shops becoming a lot more budget aware and preferring near-shore hiring to save costs while still getting lots of really talented devs working on the product.

Yeah, I think the big shift to remote-first work has played havoc with salaries (in a bad way for some people but in a good way for most people).

https://www.crunchbase.com/organization/tigerbeetle

Nubank which pays a salary based on the local Brazil market and also being one of the biggest Clojure shops, these two facts can definitely contribute for the decrease of these numbers

I know clojure is being accepted into traditional banks in Singapore as well

I think averages hide a lot of information - it's hard to know for sure without seeing the actual distribution of the salaries. Three things that may have happened: • Either more jobs got added below the previous average figure • Some jobs got lost above the previous average figure • Both of the above.

I know the tech press gets a lot of milage out of the Stack Overflow survey but I think it's a mistake to read too much into an online survey which is entirely self-selected. It's like asking whether people get a better salary if the Harry Potter character who is most like them is Ron or Hermione