Paying For Free Stuff Is a Solved Problem, Isn't It?

A couple of days ago, the LHC Blog asked about the future funding of the arxiv pre-print server, currently hosted at Cornell. Cornell is looking to get some external funding, though:

Currently the plan is to ask the "heaviest user institutions" (other university library systems) to voluntarily contribute to support arXiv operational costs. The FAQ states that the library has already secured commitments from 11 of the 20 institutions that make the most use of the arXiv. (I've seen an unofficial list; these include many of the 'big name research institutes' around the world.) In return, besides academic karma, these institutions will be recognized for their support with arXiv banners and would possibly be privy to more detailed arXiv usage statistics. The target appears to have such contributions support a fraction of the operating budget.

I dunno. I would've thought the solution was obvious. After all, countless blogs and websites give away their content for free-- this is a solved problem. All you need to do is to add Google ads to the arxiv template. That way, everybody looking for physics pre-prints can see the same pseudo-relevant stuff I get every time I read a message in GMail...

Given the amount of traffic the arxiv gets, they ought to be able to get a pretty good rate. The real question is whether they should charge a flat rate per ad, or use a sliding scale based on the authors (increase by a factor of two for each Nobel laureate in the author list, etc.). They're smart people, though. I'm sure they'll be able to find the most effective way to allocate the ad space...

More like this

I don't agree. I think we should make the effort, if at all possible, to keep the core of academia advertising free. I don't want to go to the McDonald's university and I don't want to see adverts for quantum flapdoodle alongside my arXiv papers.

The annoying thing is that, at a $400k budget, it's expensive by the standards of university library expenses but would be an utterly trivial expense for the NSF. But they don't want to fund its daily operations, because it isn't "research" -- even though it plays a crucial role in the research of hundreds or thousands of people.

I don't agree. I think we should make the effort, if at all possible, to keep the core of academia advertising free.

I wasn't entirely serious.

The annoying thing is that, at a $400k budget, it's expensive by the standards of university library expenses but would be an utterly trivial expense for the NSF.

Well, it would be three fewer research grants, more or less, based on the average size of the proposals the NSF sends me to read (typically around $400K for three years). Do you want to be the guy whose funding gets denied to fund the arxiv?

A perhaps relevant PhD

(And by god it's sad that I remember this nearly 4 years later)

"I don't agree. I think we should make the effort, if at all possible, to keep the core of academia advertising free."

Well, a number of journals have advertising, including the very largest ones. Open an issue of Science or Nature and you'll get hit with full-page spreads of ethnically diverse models in white coats vaguely waving pipettes in the general direction of some new biomedical whiz-bang apparatus.

Why advertising is fine in print but not online escapes me.

What we ought to do is take all the various money that goes into NSF grants for publication fees, that professional societies spend to support journals, and funnel that to ArXiV.org....

OK, journals aren't completely obsolete. We still need peer-review. MAYBE we need copy editing, although as somebody who's been not at a university for two years and thus getting my papers only from arxiv.org, I'm not sure that even that has enough value in it to be worth the money we spend on it.

For peer review, all we really need is to pay a couple of editors to manage it, since the actual review is done by scientists without any payment from the journal. Perhaps we also pay copy editors. But we should stop paying for all the layout, and stop paying various journals to claim copyrights on and then lock up the "official published" content for a year or two (if it's a real journal) or forever (if it's *Nature*)... sending many of us to arxiv.org to get even fully published articles.

What I'm trying to say is that we already have systems in place to pay for publications, and we're not using it very effectively given the modern era.

Janne, the problem is that advertising can be a slippery slope. Do you want journal editors to be thinking about choosing content that can drive ad revenue (e.g., lots of controversy = lots of hits/issues sold), or do you want them thinking about scientific merit? Not to knock the glossy journals, but surely they are aware that they're trying to maximize revenue, at least in the for-profit ones.

Doug, I don't doubt the risk of a slippery slope, or of thinking profit before content. What I wanted to point out was that to the extent there is a problem we already have it in paper journals, so advertising online would not be anything new or different.

The part I don't understand is that serving stuff for free (or at least really, really cheap) is also pretty well solved. I'm not that familiar with the inner workings of arxiv, but what the heck are they spending the $400K on? Does it really need that much staff? It sure can't be for hardware or bandwidth.

$400K/year doesn't sound like that much to me. You can blow $250-300k/year on that on two people to maintain the site, if you pay them what they're worth and include benefits. And, I suspect that arxiv.org uses a *lot* of bandwidth.