Jamendo’s artist donations are paltry

Torrentfreak’s Ernesto crunched the artist donation numbers on Jamendo, the awesome Creative-Commons powered free music community:

Of the 423968 users, 1650 have donated something, little under 0.5%. In total, these users were good for 2712 donations adding up to just over $36,000. This translates into an average of little over $10 per donation. The largest donation on Jamendo thus far was 200 Euros ($250) [...] Jamendo currently has close to 10,000 artists (not all of them accept donations), and 648 of those received at least one donation.

(If you doubt Ernesto’s figures, see for yourself). Even factoring in Jamendo’s sharing of advertising revenue with artists, it’s pretty clear that there are not many indie musicians out there supporting themselves through Jamendo donations. It’s hard to draw general conclusions about this future of the music industry as this sample is heavily biased: Jamendo probably attracts more serious music lovers, who donate more than the average music consumer does or would, while there is large volume of utter crap on the site which never had much hope of drawing a profit, deflating the average donation per artist.

Nevertheless, I take this as weak confirmation of my belief that donations will never significantly replace old-media sales as a major pillar of artist profits. As Enernesto points out, a lot of minor and\or independent artists are realistic about this and choose to release their music through Jamendo or similar free channels in order to capture other benefits such as acclaim, merchandise sales and concert attendance. I’m not all that confident about the merchandise-and-concerts model either: there are only so many concerts even a die-hard music lover can go to. Many people have tried to convince me that only a handful of megastar artists were scraping massive profits from music sales anyway, and the inevitable (and largely complete) shift to free media will not affect the living wages of indie artists for whom it has always been a labour of love. I’ll believe it when I see the statistics.

In related news, I was intrigued by the decision of Magnatune, a creative-commons music store, to switch to a pay-what-you want model for their DRM-free all-you-can-eat subscription. I particularly loved the subtle bit of behavioural framing below the text box for the customer’s chosen price. It will be interesting to see how it pans out, although as far as I know Magnatune is not particularly open when it comes to sales figures. I was tempted to sign up myself, until I remembered that the limiting factor on my own Jamendo addiction was not catalogue range or quality, but time and bandwidth.

Great insights about the internet and the Coase theorem in this week’s EconTalk

Ronald Coase is easily my favorite economist, and one of my favorite thinkers in general. Among his many insights was the idea of “transaction costs”, or the cost of an economic transaction which may or may not be higher than the benefit accrued from that transaction.

Coase’s classic example was the firm. Under standard models of perfect competition, he pointed out, firms have no reason to exist. If the free market is always on average more efficient than centralised, command-and-control allocation of tasks and the means of production, it should be cheaper for an employer to contract out for a service rather than hire and organise a group of people to do it. Yet it is clear that a world where every internal task of a firm, from ordering to the mail room to payroll management, had to be accomplished by a tangle of individual contracts would rapidly grind to a halt.

The key, said Coase, are the transaction costs: the cost of organising each individual transaction with a private contractor is greater than the cost of hiring somebody to do it for you, even if hiring may result in a slightly less efficient worker. Coase also pointed out that there is a maximum limit where the decreased efficiency cost begins to exceed the transaction cost, which is why firms still outsource a lot of their activities. It is not efficient, for example, for a private school to run the graphite mine to make its pencils.

In this week’s episode of the EconTalk podcast, internet and organisational guru Clay Shirky describes how the internet is lowering Coasean transaction costs for many exchanges, to the point where ‘products’ that were previously impossible due to the cost of organisation can emerge. The value of this podcast is more in the lucid examples than in the insights, which are not entirely original. I particularly enjoyed Shirky’s description of the process by which Flickr co-ordinates individual activities to produce collections of photography which would have been prohibitively expensive in the pre-internet era. His analysis of Wikipedia also goes deeper than the usual vague references to ‘crowd-sourcing’, and he cleverly uses the unexpected success of voluntary, collaborative activities such as many open source software projects to point out the importance of new behavioural economics insights. Unfortunately, the podcast ends with a strange and poorly explained hope that creative commons-like vehicles can be used to better organise collective action in the real world. Isn’t that where the high transaction costs we are trying to escape were to begin with?

Australia to filter internet access

I made a conscious decision after the last federal election to stop following Australian politics. It took me a little longer than most Aussie children to realise that, like Australian history and Australian film, Australian politics is largely dumb, boring and irrelevant to daily life. So it was a little surprising to read today that our benevolent rulers plan to force ISPs to filter ‘illegal’ material from all internet traffic, and to block pornography and other ‘inappropriate content’ from users who do not specifically request an opt-out. This is idiotic for so many reasons I don’t even know where to start.

Firstly, let’s look at the practical aspects. As anyone who has come into contact with content filtering software at school or in a home knows, there is no way to filter internet traffic with any reasonable accuracy. Rivers of ‘objectionable’ data are going to get through, and lakes of perfectly unobjectionable content will be dammed behind insensate walls. If this ridiculous plan is actually implemented, this will result in incredible pressure on the government to ‘do something!’ about all the nasty stuff still getting through, and to ‘do something!’ about all the good, important stuff still getting blocked, which will drive the government to ‘do something!’ to make the filter more complicated, more expensive and a bigger speed hump for Australia’s already sluggish internet access. To top it off, for anybody who actually wants access to ‘objectionable’ material, there is no shortage of workarounds ranging from simple browser extensions to the old-fashioned methods of bringing smut across national borders. According to the government’s own report, none of the candidate filters work on non-web protocols such as P2P or instant messaging networks. In the worst case scenario, the government will thus begin cracking down on circumvention technologies themselves – and we all know how well that’s worked in the past. In short, any ‘great firewall’ will be both ineffective for its intended purpose, and a hassle to everyone regardless.

Second come the reasons for having such a filter at all. According to the current plan, there will be two blacklists: one covering illegal material, which will be universally applied, and one covering ‘not child-friendly’ material, which will be opt-out. What possible justification can be given for this second filter other than the most blatant, cynical populism? Anybody who wants to filter their children’s internet access can go out and buy an inaccurate, easily circumvented filter to install in their own homes. Why do we need an inaccurate, easily circumvented filter for every ISP? What if parents want more fine-grained control over their filtering? Who takes the blame when parents rely on centralized filters, and something slips though the net? What poor souls at the ISP will be forced to field tens of thousands of tech support calls from confused, angry parents? How is will the opt-out be administrated anyway? Will the government subsidise ISPs for the administration costs involved?

When it comes to the illegal blacklist, the ramifications do not need to be spelled out. I’ll simply say that anybody who thinks the chance of a ’slippery slope’ into Orwellian state censorship of thought is overblown should note that this is exactly what is being explicitly proposed. No slippery slope is required. Sedition is illegal under Australian law; the filter, as described, would universally block access to material critical of the government. Information about other illegal activities, such as drug use or euthanasia, could also be blocked. It would be naive to think the filter will meticulously remove all content of dubious legality; given that it is a purely populist proposition, it would be a counterproductive to introduce any drastically unpopular restrictions. On the fringes, however – and to many people sedition, drug use and euthanasia are decidedly fringe topics -  there is plenty of room for incredible restrictions on free speech which would go largely unchallenged by a majority of the population.

For what it’s worth, should the great firewall ever come to fruition, I pledge to post any and all circumvention methods I hear about to this blog. Let’s hope that it never has to come to that.

ADDED: Check out this blog, this activism site and this amusing parody site.

I’m putting in the poll below purely to try out the new WordPress polls thing. Have fun with it.

Children are not second-class humans

What’s wrong with this picture? (Apart from the over-repetition of a phrase I would be hypocritical to condemn).

It took me a little while to figure out why it bothered me so much. Take a look at the list of possible ‘policy violations’ and ‘inappropriate behavior’: no chewing gum or public displays of affection in this school. Who gets to decide what language is inappropriate, or what behavior is disrespectful or disruptive? Could not accepting “no” lead to – never! – independent thought?

I’m not insensible to the fact that children will generally grow up to be ratbags unless given boundaries while growing up. In fact, I would go as far as to say that parents have a duty to moderate the behavior of their children in the same way that they have a duty to provide them with food or housing. Yet this is too often taken to mean that children are a second-class kind of human, not yet ready for the rights of self-ownership and freedom magically endowed upon them us the second we turn eighteen. If parents have a duty to limit and guide their children, this unconscious logic goes, surely it is reasonable that they have the right to do so.

This ‘duty to infringe a right’ is true, as far as it goes. Unfortunately, the way our societies treat children goes far beyond the license allowed by this reasonable position, yet we are so used to seeing children as second-class humans that we completely fail to recognize it. We take it for granted that schools are authoritarian cages, because we know that without discipline most schools would be madhouses. Yet the need for discipline is a specific reason to allow for specific limits on the freedom of children; to assume that these specific cases excuse authoritarianism in all aspects of the life of a child is a case of faulty generalisation. The default status of children, as it is of all humans in the libertarian conception, is total self-ownership and freedom. Reasonable cases can be made for encroaching on these freedoms, but these must be argued one at a time and never taken for granted.

This is especially true when children are compelled to go to school. In adult life, we accept limits on our behavior when we are on or using other people’s property. Although we are free to smoke, we accept that a shopkeeper has the right to throw us out of the store if we light up against their wishes. Thus, the imposition of rules on in schools seems doubly legitimate: children need to have their behavior limited, and the teachers have the right to enforce standards of behavior on school grounds. Yet children are compelled to go to school, and rarely even have a choice in which school to attend. They are forced to spend a significant portion of their waking lives on somebody else’s property, and therefore to follow somebody else’s rules. If an adult was kidnapped and held by force on somebody else’s property, we would not consider it illegitimate for them to flout the property owner’s rules.

Again, this is not to say that all school discipline is illegitimate. I am simply pointing out that the criterion by which it can be justified is narrow and specific – namely, that the encroachment on the child’s freedom is part of the parent’s, and by extension the teacher’s, duty of care – rather than broad and unrestricted. The onus is on the adult world to demonstrate, in each and every case, why a right should be taken away. This includes an onus to demonstrate that there is not a better, more free way of achieving the same end. Libertarian paternalism tries to increase freedoms in the adult world by influencing rather than controlling behavior. Yet the idea of ’soft’ behavior modification often seems absurd in the school context, and not infrequently in parenting: we argue out loud that children will run wild given half the chance; and believe subconsciously that there is nothing really wrong with encroaching on their freedom anyway.

The second point I have already delt with. The first is, again, quite reasonable on the surface. Children get up to enough trouble in school and at home as it is; it seems natural to assume that, if the boundaries are lifted, their behavior would become correspondingly worse. Again, this is a faulty generalisation that breaks down when we look at specific cases.

Let’s go back to the detention note linked to at the beginning of this post. Assuming that the school is in the US, the code of ‘uniform violations’ is probably targeted at crude or bigoted slogans or images. (Most uniform codes at non-US high schools, such as the one I went to here in Sydney, cover such heinous sins as ’sleeves too long’ and is are not worth even trying to defend). Since dictating what a person can and cannot display on their clothes is a violation of their freedom of speech and expression, there must surely be a correspondingly weighty benefit to the child’s upbringing of this policy. It can’t be reasonably argued that this benefit accrues to the child wearing the clothing: they would have seen it anyway, so it’s not protecting them from any unsavoury ideas; and they are free to wear it outside school, so it’s not modifying their behaviour in any real way. Is it to protect other children from crude language, sexual content or so on? Anybody who thinks that children are not exposed to sex, drugs, violence, bigotry or crude language and could thus have their innocence shattered by a t-shirt slogan is a fool with a very poor memory of their own childhood. Is it out of fear that it will provoke unruly behavior in the other students? This is possible, yet unlikely; and it is a very thin premise on which to restrict so fundamental a freedom.

The most reasonable explanation for uniform policies, and ironically the one generally given to students in Australian schools, is to maintain appearances. School administrators have strong incentives to find simple ways to cause big gains in the perceived quality of their schools: private schools have money on the line, and public school administrators have their reputations, performance-linked pay incentives and ultimately their jobs. Enforcing a uniform policy is a cheap and effective way to present a more polished image. It costs only the freedom of students, which nobody cares about, and the cost of enforcement, which is marginal and accrues mainly to teachers. Absolute power corrupts absolutely, and when a school administrator is given near-absolute power over their students, it is natural that they will frame rules in their own interest.

Not all school rules are of this type. Many are reasonable restrictions on behaviour which generally do advance the care of the child. Others, like ‘no littering’, are reasonable compromises between a school administrator compelled to maintain the property over which they hold stewardship and the compulsion of students to attend in the first place. Nobody has ever mastered the art of parenting or teaching, and there will always be difficulty in establishing what is genuinely worth forcing children to do. In general, however, the socially accepted level of control lies far, far over the line into unjustified authoritarianism. The rule against ‘public displays of affection’ (PDA) may have had the wrong-headed intention of preventing promiscuity among teenagers, but in reality simply curbs their freedom of association and social interaction while pushing out of sight what they’re going to do regardless. If you want a reality check when considering a rule imposed on children, consider the same rule being imposed on yourself, or upon an adult if you are a child. How would an adult react if forced to spend 6 hours a day on somebody else’s property, and then told that kissing their partner during that time was forbidden? It is not exaggeration to say that this is the kind of authoritarianism for which the west routinely condemns Islamic states, yet it somehow it can be imposed upon children without a hint of outrage. Many adults may find (or pretend to find) the idea of teenage sexuality repugnant, yet repugnance has never given anyone the right to tell another what to do.

Children are not second class citizens. They are humans with the full rights and freedoms naturally accorded to us all. The only difference between a child and an adult is that children are owed a duty of care from their parents, and that when well justified, this duty can extend to putting boundaries on the child’s freedom. These boundaries, however, must be constantly justified, and the least freedom-infringing methods constantly sought. This doesn’t come naturally. The second-class position of children is ingrained into our culture and our institutions. This does not, however, make it right. Worse, children do not have anything like the voice that adults do in defending their rights. Power corrupts, and adults routinely exploit this defenselessness to take what rights from children they can. This does not make it legitimate, and it does not make it right.

“Comes with music” will fail

This week’s edition of the Economist newspaper features an article on “Comes with music” (CWM), the latest addition to the legion of subscription-based “solutions” to the file-sharing “problem”. Under CWM, mobile phone makers purchase 12 month subscriptions to online-music catalogues and bundle them with handsets. The idea is that end users will not mind paying for music if the cost is hidden inside the price of something more tangible. CWM is touted as a win-win-win model: handset makers get an extra feature to entice customers, end users get “free” music, and record companies get to make a profit. The Economist describes CWM as “potentially a big step forward”. I think it is a half-baked idea which is doomed to failure.

The first reason is the restricted nature of the service. Major record companies are a lot like alcohol addicts: although they know as well as anyone how much it hurts them, they just can’t compel themselves to give up a bad habit. So it is no surprise that the tracks available on CWM are strait-jacketed with Digital Rights Managment (DRM), the asinine technology which attempts to restrict users from moving or playing music files in anything but the most strictly prescribed way. Of course, even if the DRM method Nokia is using remains uncracked, nobody is going to put up with the restrictions or pay to have them lifted when they can just download the music from a P2P file-sharing network, unrestricted and free. At best, all DRM will achieve is to encourage users to use CWM when convenient – when they want a certain song, right now, on their mobile handset – and carry on downloading the bulk of their music the normal, illegal way. On top of this, the twelve month expiry limit is simply irrelevant to a generation for whom downloading is the norm. If the users even notice that CWM has expired, they will hardly care.

The second reason is that, of course, CWM is not free. The handset maker has to pay at least one record company for the subscription, and that cost ultimately must get passed on to the consumer. Nokia may choose to cross-subsidise the service with profits from other handsets, or cut other features from CWM handsets, or whatever. Regardless, handsets with CWM will be at a competitive disadvantage against handsets without CWM. This is only an intuition, but I strongly doubt that the addition of CWM will sway consumers purchasing decisions more than an infinitesimal amount. It simply isn’t a shiny, attractive feature when file-sharing is so easy and familiar, and when mobile carriers and ISPs are already pushing so many other DRM-stunted, mutually incompatible digital music services, subscription and otherwise. The Economist article even suggested that the “unlimited” downloads will in fact be capped by a “fair use” limit, which will tarnish CWM’s image yet further. Unless Nokia can find a way to make CWM sexy – a very tall order – or to absorb the cost, CWM handsets will not be competitive.

The final and most basic reason is that the music industry has left it far too late to supply an alternative to file-sharing. My generation grew up with Napster, Limewire and BitTorrent the way our parents grew up with television, air conditioning and jumbo jets. For us, file-sharing is not a way to get music, it’s the way. A technologically disinclined friend recently bought a new laptop and asked me what software he should put on it. I suggested, among other things, OpenOffice. He recoiled: even though OO’s user interface is a near clone of MS Office, and OO is entirely free, he would rather find a pirated copy of MS Office than deal with a slightly unfamiliar interface. He also turned down my suggestion of Vuze as his file-sharing client; he had been using Limewire for years and didn’t want to learn a new system. There are a lot of people like my friend out there, and a lot more who are open to new technologies but who just like file-sharing better. Why would they bother adopting a restricted, closed-catalogue, closed-medium system when they could just keep on downloading like they always have? Boosters of CWM must remember that it is competing against a file-sharing networks which are free, ubiquitous, near perfectly stocked and completely familiar. It’s hard to imagine a better music delivery system, much less one that would turn a profit.

I don’t know how musicians are going to make a living as music sales continue to plummet, but CWM is not part of the solution. My strong intuition is that anything which requires users to pay for music (or any digital content) is a lost cause, and profits will have to come through adjunct channels. However, I’m skeptical that donations, concert tickets, merchandise sales or advertising will entirely cover the gap. It will be fascinating to see what solutions the market comes up with. At least we can be assured that great music can be given away sustainably.

Image credit: “stereo geisha white“, by chotda, under a Creative Commons Attribution-Noncommercial-No Derivative Works 2.0 Generic license.

Funky french ska album, released under CC

I’ve been compulsively re-listening to the short but sweet 2007 latin/ska/rock album Kasane! by French band INTI. Like all albums on Jamendo, it’s released under a creative commons license, meaning you can download it for free, completely legal and legit.

You can download the album from Jamendo, or spare their servers and use one of these torrents: MP3 (zipped) or OGG (zipped).

Mini-review: Bibus, an open source bibliographic managment tool

After converting from Ubuntu linux about a year ago, I found that there were only two tasks that required me to boot into my Windows installation: syncing my iPod touch with iTunes, and writing documents that required heavy scientific referencing with EndNote. The first task may very soon have a viable linux alternative with the imminent release of PwnPlayer. When I recently set to work on a lengthy literature review, I decided to go and find a viable alternative for referencing too. After testing a few options, I came across Bibus. It’s not perfect, but it does competently replace all the functions of EndNote – and, of course, it’s totally free and open source.

Bibus performs all the basic referencing functions: it imports citations from files or manual input, stores and organises them in a simple database, inserts in-text citations into OpenOffice documents (MS Word is apparently also supported), and updates a formatted reference list at the end of the document. It was exceptionally stable, with not a single crash, and installation, database creation and integration into OpenOffice was streamlined and easy. There were a number of little usability hitches, however, which still require some polishing:

• Importing from Google Scholar, where I get the majority of my references, is not very streamlined. I found the best way was to open an ‘Import text window’ in Bibus, and set my Scholar preferences to offer a text-only BibTeX citation for each reference in the search results. I then had to copy and paste the citation into the inport window and hit ‘import’ and drag it from the import buffer into the reference list proper. The need for this buffer was not obvious and it felt like a totally unnessecarilay extra step.
• Inserting the reference into OpenOffice was also a little awkward. EndNote adds a toolbar which lets you insert an in-text citation from inside a MS Word window. Bibus makes you mark the spot in the OpenOffice document, change windows into Bibus and insert, then go back to OpenOffice to make sure it inserted correctly. This is a minor niggle, but can get tedious when you are inserting multiple citations for the same reference.
• The way Bibus imported the BibTeX data was a little haphazard, with it occasionally keeping curly brackets “{“, the BibTeX field delimiters, in the text of the reference itself. I couldn’t tell if this was an error on Bibus or Google Scholar’s side, but either way it meant I had to manually go though at the end and remove a number of stray curly braces from my references section.
• Formatting of the references was inconsistant, even when I repetedly applied a single ruleset to the entire list. Again, this could have been a problem on Scholar’s side.