Hacker News Reader

Comment by pwthornton | Hacker News Reader

One of the reasons you are probably seeing this happen are budget cuts. Major university systems have seen their funding be cut by state governments. Elsevier charges an obscene amount of money to access information that is often publicly funded.

If money were no object, you'd probably see less university systems rejecting Elsevier. But money is becoming a bigger issue.

Ultimately, this is starting to put major university systems in line with individual users, and we should see an explosion of open-access information in the next decade.

Remember when Harvard University said it can't afford journal publishers' price: https://www.theguardian.com/science/2012/apr/24/harvard-univ...

View on HN

Well... the other reason might be that the German Universities got together to negotiate with the big publishers. While agreements are in reach or have been reached with most of them, Elsevier continued to try to throw it's weight around leading to most German academic institutions letting the subscription to Elsevier journals lapse (or cancelled) at the end of 2017, with the remaining important ones following suite at the end of 2018.

Currently almost no academic institution in Germany is subscribed to Elsevier. Here is the full list of German Universities and Labs currently not subscribed to Elsevier:

https://www.projekt-deal.de/vertragskundigungen-elsevier-201...

Elsevier did not actually shut down access to its journals for German institutions.

Wiley on the other hand has just signed a contract with DEAL, and negotiations with Springer Nature are going well enough so interim contracts are in place until they are concluded.

https://www.projekt-deal.de/

Actually, the overall budgets of universities have increased substantially— they've more than made up for the lost state funding by raising tuition and other fundraising.

U.C. Berkeley, for instance, has increased its overall operating budget by from $1.6B in 2006 to $2.7B in 2018. That's a 68% increase over just the last 12 years! [1]

Universities are spending, and taking in, more than ever.

Even with all this cash, Elsevier's subscriptions are not cheap. The estimated cost for U.C. Berkeley is $1.15M per year. [2]

[1] https://controller.berkeley.edu/home/uc-berkeley-financial-r...

[2] https://www.quora.com/What-is-the-cost-for-a-university-subs...

Just for reference, and this does not diminish the parent's point significantly, inflation accounts for approximately 24% of that increase. That's still a 36% real increase in operating budget assuming student size held constant.

> assuming student size held constant

It did not: https://pages.github.berkeley.edu/OPA/our-berkeley/enroll-hi... During this period, the undergraduate student population grew about 20%.

I was at Berkeley during most of this period. The undergraduate population swelled at the same time that the university was dealing with difficult budget cuts. The growth was significant enough that many students had, and still have, difficulty signing up for the courses they needed.

(I believe the cuts and the student population increase were related: admitting more students, especially out of state students, brought in more revenue. But there could have been other factors, like more students wanting a college degree after the financial crisis.)

To color this even further, at UC Berkeley I had intro courses with 1,200 students.

Perhaps, and budget cuts can justify more risk taking. No one wants to be accused of taking political stances that go against student interests, but budget cuts are budget cuts.

In my field, where being Free Software based is a selling argument, I like to joke that recessions are good for our field, and economic booms are good for our field (https://www.youtube.com/watch?v=pgUA1tluVmE). I just adjust my pitch depending on the circumstances (save money or invest in a more powerful platform).

The other reason is of course Scihub. You pretty much can't do research without access to Elsevier's journal, so before Scihub, university libraries just didn't have a choice. Now they do.

Of course, none of them are ever going to say in a million years "Yeah our researchers don't mind so much now that they can get any paper illegally. Extortion over.", but it doesn't take much reading between the lines.

> Of course, none of them are ever going to say in a million years "Yeah our researchers don't mind so much now that they can get any paper illegally. Extortion over.

Not only do universities not openly admit to SciHub usage, but there are even university administrations that still warn their students and staff against using SciHub.

I recently read through a fairly prominent European university’s freshly-revised course for undergraduates on how basic use of the library and research methods (I saw this as part of my job, so naturally I won’t name the institution). Among the course content on how to find journal articles in JSTOR and Web of Science, how to properly cite your sources, etc. were admonitions not to use SciHub because it “is illegal”, and warnings that use of SciHub can lead to failing a course or having one’s student status revoked.

Parallel construction it is, then.

Cite the official published source.

I wonder how this dichotomy of "use Wikipedia and SciHub, but you don't cite them directly" affects the ethics of graduates.

Citing Wikipedia directly is silly when Wikipedia often has its own citation of a primary source that you can use instead. Wikipedia is explicitly not a primary source by design.

SciHub is different because it replicates primary sources.

I'd argue that it's unethical to keep publicly funded science behind a paywall, so perhaps the ethical question you're implying is more nuanced than that of simple "piracy".

People cite Wikipedia all the time in peer-reviewed journal papers. Many e.g. high school teachers recommend that their students not cite (any) encyclopedias directly, because they want the students to learn how to read and cite primary sources.

Reading a secondary source for an overview then diving into primary sources is a perfectly fine and ethical way to do research. I can’t see how schools and teachers advising students to read secondary sources for context but then follow their references to primary sources to read for the details would cause any kind of ethical compromise.

There is no reason to cite SciHub (just like there is no reason to cite Academia.edu, ResearchGate, some professor’s personal webpage with a hosted preprint, Google Scholar, etc.) whether or not someone uses it.

Sci-Hub isn't a thing that you cite. It gives you a publication from a conference / journal. That's what you cite. Citing Sci-Hub would be like citing google.

I think you have a point and a lot of replies are missing it.

Tacit acceptance of Sci-Hub is tacit acceptance that some laws don't matter and everyone knows it. That's certainly an ethical stance with implications. It's not necessarily a bad stance (D&D would call it "chaotic good") and it is I think my own stance, but once you decide that laws are secondary to your personal and professional needs if you don't get caught, you're likely to use the same argument in other respects. Citing the original paper instead of the Sci-Hub URL only increases the extent to which you're pretending to follow the law.

Wikipedia has primary sources linked, yes, but it has its own biases. It tries not to, but it still has them to some extent, perhaps just in terms of what sections have been written and what haven't based on editors' interests. If you use Wikipedia to find primary sources, you're filtering your research through those biases. That has direct implications, and indirect implications for ethics in that you've decided this filtering is acceptable to you and also that it's acceptable (and perhaps required) not to acknowledge it. (And yes, of course there are other biases in terms of what university libraries stock, what previous academics have written about, what funding agencies are interested in, etc. I'd say that how much you pay attention to these potential biases and whether you see them as significant is also a question of ethics.)

How many grad students for generations have been smoking dope despite the laws against it? Using sci-hub despite the law is not a significant departure from the status quo of grad students ignoring dumb laws when it suits them. Sci-hub does not present a uniquely deleterious effect on the ethics of grad students.

Also, even if sci-hub were legal, you still wouldn't cite the sci-hub url. That's not how things are cited.

> once you decide that laws are secondary to your personal and professional needs if you don't get caught, you're likely to use the same argument in other respect

I don't think it's that clear cut. If you drive above the speed limit, or smoke weed in a state where it's not allowed, or pirate some DVDs, you're not automatically going to progress to more serious crimes. People seem to be perfectly capable of "ahh let's ignore this stupid law" and not start sliding down some slippery slope.

Because some laws just don't matter. This is a good thing to realize for people. Laws aren't automatically right and moral just because they're laws. There are plenty of reasons why a law can exist, and morality usually isn't the reason.

> but once you decide that laws are secondary to your personal and professional needs if you don't get caught, you're likely to use the same argument in other respects.

I think it's more nuanced than this. Something being illegal and immoral, are two different things (there's overlap, but not always). Some laws are completely wrong and breaking them is not immoral. IMO (you may disagree) academic articles "belonging" to the publishers and being copyrighted, is one of them. Hence, I'd rephrase what you wrote above as: "but once you decide that laws are secondary to your [...] needs [...], if they're terrible laws that should be changed, (but probably won't be, at least in the near future, due to inertia and lobbying), you're likely to use the same argument in other respects", which is much less problematic and in some ways a good thing.

Obviously, motivated reasoning is an issue, so you should be very careful not to fall into that trap and start breaking laws willy-nilly arguing that they're all bad laws.

> Citing the original paper instead of the Sci-Hub URL only increases the extent to which you're pretending to follow the law.

How is it at all relevant how I got a paper? Once upon a time, there were paper copies of journals in libraries. If you were citing a paper at that time, you wouldn't write "Journal of X, found on the 3rd shelf, in the 5th aisle of the Y library". In current times, you can (usually) legally get a paper by pestering the original author for a copy, by e-mail. If you do, you wouldn't cite the paper as "obtained from Original Researches, mailto:zzz"). (If you put a "link" in the citation, it should be the DOI (digital object identifier) as it's the universal(-ish), immutable (hopefully) identifier.)

Nobody needs to cite scihub. Download the papers you need off of it, cite them.

Citations should always be in some standard form, including author, year, paper title, journal. But for a web link, just put the DOI. A reader can either click it or paste it into scihub, as they prefer.

There is a lot of money going from universities to publishers, be them traditional or Gold OA. Researchers would rather use either Green OA (like arXiv) or Sci-Hub or articles available on other researchers pages.

Is therefore interesting that a big university makes this move, with the argument: "the publisher would have charged UC authors large publishing fees on top of the university’s multi-million dollar subscription"

Yeah it's very much a don't ask, don't tell sort of thing. Check out some of the location images for scihub users and they very much paint a different story.

From that information, I would actively want to check SciHub out if I had never heard of it. Academia + notoriety sounds novel and useful.

Part of why SciHub has succeeded is just how easy it is. Copy/pasting an arbitrary DOI from any source will always beat most journals arcane login systems (especially on mobile) and their divine mission of obscuring "Download PDF" buttons.

SciHub is amazing. I'm not in academia or industry, and I was working on trying to find some eutectics of simple ionic halide mixtures with some software that was missing some of the activation energy coefficients for some of the ion interactions. Must have looked at at least 30+ papers, which if university-publishing house cartel have their way, would have cost me thousands of dollars just to look at. Between sci-hub and all the open source code from labs and individuals around the world, it's just breathtaking what can be explored now outside the confines of academia.

Curious what the application was?

Mostly exploring molten salt mixtures for thermal energy storage with a larger working temperature range and a reasonable melting point compared to what's used typically.

Do you work at a molten salt energy storage company? What's the industry landscape there like? Like still primarily R&D, or is it mostly about production now.

No I don't, just a renewable energy hobbyist since about a decade ago when I interned at P&G chemicals researching transesterifcation/industrial processes for biodiesel production. I mostly want to experiment my own micro CSP system, but never really had to tools to explore salt combinations theoretically in a way that was fun for me.

Warning people not to use sci-hub, mentioning it by name as something that you shouldn't use, is a plausibly deniable way to spread the word about it.

Oh, I can assure you that this particular university is dead serious about SciHub use being undesirable. The course I mentioned dedicates significant space to respecting copyright, not only as something expected of researchers, but as a moral duty.

Also, inter-library loan is a source of income for the university library, that is, if the library does not have access to a particular journal in the databases it subscribes to, then it will obtain the journal article from elsewhere but at a hefty fee, so perhaps this is another reason they would prefer students avoid SciHub.

Enlightened morality often means being mindful of tradeoffs.

Not doing an exhaustive search of academic literature because it is prohibitively expensive or time-consuming is a clear negative -- doing so may result in lower-quality research.

I personally see many good reasons to distinguish general intellectual property theft from using SciHub for academic research.

Which university is this? If anything involving copyright is "moral," it's breaking it.

Who defines "moral"? I can't see that breaking any law is the moral thing to do as the law is the moral code that society agreed to give it self (at least in a democracy).

By your logic it might become moral to murder, because it benefits you and maybe many others more than the one murdered.

Following the law is always a moral obligation and personally deciding which rules should be followed and which shouldn't is a dangerous space. Especially from a biased and self-centered world view, which we all have as individuals.

Or, you know, it's the law corporations were able to push into existence through money. Law isn't necessarily a reflection of our moral code. It can be. Parts of it are (Grundgesetz or the Constitution, sure), but plenty of laws aren't decided by morality and aren't the arbiters of it either and even if they reflect the morality of some people, it isn't morally justified. Or people wouldn't have ever fought for black rights (Jim Crow laws say hello).

> "Who defines "moral"?"

You do. And I do. You define it for you, and I define it for me. You and I are both going to define murder as immoral, and not commit murder. Should one of us define it otherwise, the other will in practice likely be powerless to prevent the outcome.

This is the way the world works. At the end of the day, there is no universal ruleset that all humans will follow like automata; we each get to make our own path. The best we can do is attempt to persuade others to change their moral framework and to gang up with people who agree with us to create systems that inhibit those who disagree with us.

There is no sense moaning about this state of affairs because this is the way it's always been and the way it always will be.

> "By your logic it might become moral to murder"

Something else to keep in mind, is that morality is likely rooted in some limited extent in biology. The basis for this assertion is that across time and cultures, one form or another of law against murder has always been popular in any human civilization you can find evidence for. Even when societies carved out formal exceptions to this rule, such as the legal witch burnings in recent European history, the human sacrifices in more ancient European history, or the human sacrifices seen in Aztec civilization (that had no plausible meaningful cultural exchange with Eurasia for many thousands of years), there still exist laws in some form against murder in general. My point with this likely controversial digression is that it's unlikely that anybody who constructs an argument for the morality of academic copyright infringement will accidentally construct a persuasive argument for casual murder. Even if you use formal logic to find a link between their perspective on copyright to casual murder, that formal argument will be irrationally rejected by nearly everybody, who has an innate predisposition towards rejecting a system of legal casual murder.

tl;dr: reductio ad fortuita occidendum is not a persuasive argument.

Perhaps we need some new entries for the Devil's Dictionary:

law 1. (archaic) An agreed rule that everyone must obey. 2. A document produced by the government that grants additional powers to the police to confiscate property and otherwise punish people and to protect the police in case some troublemaker complains to a court.

lawful Of or pertaining to an act of killing committed by the government, the police, or the armed forces.

You are putting yourself in a logical trap here, associating copyright law with murder...

Which statement is true...

"By locking knowledge behind pathways you are both directly and indirectly causes the death of others by limiting information to people that could use it to save others"

"By not respecting copyright you are breaking the institutions that create scientific knowledge and making it riskier for said institutions to invest and create new things which can be used to save others"

The law is always at moral odds with itself. It must try to achieve both goals which is an impossibility.

On the other hand I have seen a guide from a Swedish university library with a "the dark side" section recommending Scihub, #icanhazpdf and /r/scholar. The contract between Elsevier and pretty much all Swedish universities was canceled a year ago and they seem to be fine.

a nods as good as a wink ;)

Oh please. Everybody I know personally and everybody I know of through these people use scihub. It's such an incredibly useful tool that it's now ubiquitous throughout academia. Of course universities aren't 'officially' advising its use, but informally, it's pretty much accepted.

How long can that last?

Actually, TFA quotes UC President Janet Napolitano basically saying that:

> "I fully support our faculty, staff and students in breaking down paywalls that hinder the sharing of groundbreaking research" ...

I wonder if UC gives any money to the Scihub admins. I found their current donation link: https://sci-hub.tw/donate

Yeah, budgets are, in part, motivating the push for open access journals. But I don't think university-wide budgets are the most relevant ones to consider.

If you look more closely at university _library_ budgets, you'll usually see that journal/e-journal/database subscriptions have been eating up bigger and bigger chunks of the pie. The ever-increasing price of a subscription to ScienceDirect.com is directly taking away money from buying monographs (books), maintaining archives and special collections, and keeping university libraries open late at night for students to study.

> when Harvard University said it can't afford journal publishers' price

Harvard's endowment is, according to Wikipedia, $39.2 billion as of FY 2018. My first reflex was to say that "can't afford" must be a relative term...

A closer look at the article says: "A memo from Harvard Library ... warning it could no longer afford the price hikes imposed by many large journal publishers, which bill the library around $3.5m a year." So I could believe that the Harvard Library is given a certain budget by the main Harvard organization, and that this budget is insufficient to support that bill plus other unavoidable recurring expenses.

You're also missing the distinction between an endowment and a revenue stream. Harvard has almost $40B in assets but they are going to be making much smaller annual returns on that — probably 10% at best — and the reason why people donate those funds is to support academic work, not boost profit margins for large companies. Harvard's academics are providing an enormous amount of unpaid work for journals so it's dubious that they are jumping to cut a big check for that rather than, say, scholarships, equipment, etc.

I was aware of that distinction while writing my comment. Even if only $30B of Harvard's assets are financial investments, and even if they earn just 5% annually from that, then that's $1.5 billion per year, so $3.5 million is just 0.23% of their income. I'm not saying $3.5 million isn't absurdly high. I'm just objecting to the statement "Harvard University [itself] ... can't afford ...", which is journalistic hyperbole.

"Under Elsevier’s proposed terms, the publisher would have charged UC authors large publishing fees on top of the university’s multi-million dollar subscription"

So it's a move against both traditional paywalled publishing and Gold OA.

The argument that some universities refuse to pay because they are "scholarly poor" [1] is only propaganda.

[1] https://chorasimilarity.wordpress.com/2019/02/28/google-tran...

> Ultimately, this is starting to put major university systems in line with individual users

So the side effect of budget cuts is actually a good thing?

Maybe, but subscriptions costs have also been rising for years, while at the same time academia has also started paying, higher and higher, publishing fees.

Maybe, and just maybe, if money was not an issue, research would also be more open.

Is there any in-deep analysis about university finances? It seems that getting around $150.000 from each new student they should have more than enough to sustain and profit.

My understanding is they are demand inelastic so their costs naturally go up, and there's been an arms race in the past 20 years to build very nice new buildings, housing, etc to compete on look, prestige and features for students. Reputation of the school is one of the biggest differentiators, so now that they're being run as a business rather than the state costs have no incentive whatsoever to decrease. Now of course, even if the states wanted to be more equitable to their citizens by going back to a more state-funded model, the cost to maintain all of those new features is massively increased. Its like an asset bubble more or less.

They are not getting $150k per student. That's the nameplate price, which is not the actual price paid by most students. The students who do pay nameplate price see much of their tuition going to fund financial aid for less well-off students.

Net tuition and fees for students at private 4-year colleges is about $14,600 in 2018-2019. That results in a total of $58k/student in revenue, not $150k. https://trends.collegeboard.org/college-pricing/figures-tabl...

As public institutions, UC campuses see even less: https://trends.collegeboard.org/college-pricing/figures-tabl...

The arms race issue is real, though, and students are less price-sensitive than you might otherwise expect because so many of them are paying for all of it with debt, and the dollar figure looks so obscene everywhere that I suspect lots of students just choose not to think about it. One school's dining hall being fancier is immediately appreciable on a college tour. One school's ridiculous high number that you won't have to pay for years vs. a slightly different ridiculous number at a different one feels less concrete, even though that difference is what paid for the fancier dining hall. Given that, there are big incentives for schools to go ahead and invest in the fancy things and win the students.

It is quite real, this reddit thread shreds some light on the issue: https://www.reddit.com/r/AskAnAmerican/comments/7qesj9/why_i...

I think it is real, but I don't think expenditures drive it all that much. What does drive it is the perceived value through pricing. If the going rate for public schools is $5k/yr, the private colleges put their nameplate rate at $25-30k/yr. Significantly more, but that higher price sends a quality signal, and the students they want don't actually pay it, so they aren't negatively impacted by the high cost.

When the state school price increases to $10k/yr or even $15k/yr, the private schools can move up their nameplate price to $40k/yr or $50k/yr and it doesn't seem ridiculous anymore, and still retains the signaling value.

There has been some increase in student services costs based on the expenditures numbers I posted (about 4%/yr real increase), but that's not enough to explain the 50-100% increases in nameplate tuition at both public and private schools. That is driven by de-investment in post-secondary education by state legislatures, which directly results in tuition increases by public schools to balance the books, and then the similar increases by private schools to maintain their relative cost.

There's plenty.

The college board has good data on the cost side: https://trends.collegeboard.org/college-pricing

And the department of education has good data on the expenditures side:

Private: https://nces.ed.gov/programs/digest/d17/tables/dt17_334.30.a...

Public: https://nces.ed.gov/programs/digest/d17/tables/dt17_334.10.a...

For one they don't get remotely close to $150,000 from each student. The average public 4-year university gets $6,100 per student per school year. That's tuition and fees minus financial aid coming from the school. It covers about half of the cost of educating students, with the other half coming from state funding and federal financial aid.

The average is $9,970 [0], for private ones is 34,740

[0] https://www.topuniversities.com/student-info/student-finance...

"Published tuition fees" is a very different figure from what the average student actually pays.

Universities in the UC system are definitely not getting $150k from each student. Tuition doesn't even cover the cost of educating each student.

You can see UCLA's financials here: https://ucla.app.box.com/v/acct-pdf-AFR-16-17

Tuition and fees netted $833MM in 2017. The cost of the undergraduate schools alone was well over a billion dollars.

Some of those "costs" are self-imposed such as administrative salaries increases, as investigated by the LATimes[0]

[0] https://www.latimes.com/local/education/la-me-uc-spending-20...

Reply on news.ycombinator.com