In math most people publish their papers on arxiv regardless of whether they get published in a journal. Arxiv has its own moderation structure, but theoretically published peer reviewed articles should be more trustworthy because they’re reviewed by your peers. In reality reviewers don’t have time to read papers super closely so some shoddy research gets through even when no blatant corruption is involved. In pure math this isn’t a huge issue because the work usually speaks for itself, but for some areas, especially applied statistics, it’s not so obvious whether an argument is actually well supported or statistically cherrypicked.

Because in a lot of hiring processes the worth of the researchers is based on this streer cred. It's a messed up system.

Even more messed up is that journals that do try to be more open about their procedures and that don't try to make a profit are marginalized or in some cases even not indexed. For example, eLife no longer has an impact factor calculated because it's experimenting with a publishing model that disincentivises profit and some other undesirable things in academic publishing.

Because the journals existed as massive, financially powerful entities. There were negotiations over open access arrangements a few years back which led to things like "gold open access", which involves papers being free to read, but costing a heckton for the researcher's in "Article Processing Charges". This happened because the journals effectively argued that "even though we're functionally useless in the modern day, and don't even provide services like copyediting or typesetting support for researchers, you can't just make research actually be fully open, because then we would no longer be able to be absurdly profitable. Won't someone think of the profits?!". And then their influence meant the open access agreements were half baked and insufficient.

However, there is a continuing movement that is pushing for actual open access — "Diamond Open Access" doesn't charge either the researchers or the readers of papers. It's still small, relatively, but it's growing, especially in the global South or amongst independent researchers who can't afford absurd Article Processing Charges. Profit driven journals have prestige on their side, but I reckon that Diamond Open Access will continue to grow as research funding becomes more scarce relative to the amount of research being done.

"The diamond model has been especially successful in Latin America-based journals (95% of OA journals[1]) following the emergence of large publicly supported platforms, such as SciELO and Redalyc. However, Diamond OA journals are under-represented in the major scholarly databases, such as Web of Science and Scopus. It is also noteworthy, that high-income countries "have the highest share of authorship in every domain and type of journal, except for diamond journals in the social sciences and humanities"."

(Source: the linked Wikipedia page)

In academia, promotion, tenure, funding, and pretty much everything someone needs to keep their job is tied to publishing in peer-reviewed journals. If I self publish I won't keep my job. If a university ran a website for publishing, they would have to reimplement the peer review process, and often there may be ~a dozen people worldwide qualified to review a particular paper so it's not just that others from within the university could review work. If a university is implementing all that, they have basically become a publisher and likely have costs they'll want to try and recoup and could foreseeably implement a fee to publish.

Don't get me wrong - the journals have a predatory and exclusive model that should be dismantled. But until we fix promotion, tenure, and funding pathways in academia that have enabled the publishers to become what they are there will always be these problems -- any other system that pops up will not get widely used because academics will be disincentivized from using it (as discussed in some other comments here).

Would be cool to see a decentralized journal or something. No idea how it would work actually work, but since research is supposed to be peer reviewed, why not just let the whole structure of it become open and decentralized?

Before cryptocurrencies really went to shit and became so toxic/scammy, I used to believe crypto could help decentralize it by democraticing the process and making it fully transparent, while also rewarding authors and peer reviewers for their work. But alas

Obligatory sci-hub plug. When I went to uni, our teachers literally told us, that the school doesn't have the money to pay for access to everything, so here's a site you can use

I too would like to know this