Menu Home End users Hacking Humour Interviews Opinons Reviews All articles Issues Books
Free Software magazine
Terry Hancock [opinions]

Some numbers on Creative Commons [creative-commons] [statistics]

The number of the current, freshly-released version of the Creative Commons licenses: 3.0. The total number of CC licensed works on the web at last estimate: 145 million. The percent of those that we would call “free”: 29%. The time it takes to double the number of free, Creative Commons licensed works: approximately 115 days.

Every so often, Creative Commons publishes some statistics intended to estimate how many people are using their licenses, as well as which licenses they use. Now that Creative Commons has just released their new, more-compatible version of their license, it seems an apt time to go over some of the success metrics for the Creative Commons licenses. Also—conveniently—I just happen to be analyzing this data for a book I’m working on, so I’m in a good position to write this.

One of the things that CC does right is to make it very easy to trace their licenses. The recommended licensing system is to link back to a unique URL for each license type and version. Both Yahoo and Google provide a means by which you can search for web pages that link to a certain URL, so it then becomes possible to estimate how many pages on the net are pointing at the CC licenses. The bulk of these, one assumes, will be licensing statements. Of course, like all data-gathering of this type, the assumptions aren’t perfect, so there is some error, but probably not enough to make the numbers violently wrong.

The CC licenses show a fairly clean exponential growth pattern. Free licenses are growing slightly faster.

The first thing that immediately jumps out at you from this plot is that there’s nothing complicated going on here: the growth is a plain exponential. The wobbles that are there are probably not that significant—they can easily be explained by changes in the methodology (such as switching from Yahoo to Google statistics).

That’s interesting, because with all the fuss going on over whether CC has “betrayed” the free culture movement by promoting “non-commercial” licenses, which has earned them a specific _non-_endorsement by Richard Stallman, and much flak from Debian and other free software organizations; with all the FUD flying from Microsoft, the RIAA, and the MPAA, and so on; you might expect to see dips and jumps in the license adoption rate. But if such political upheavals have any effect, it must be pretty small, because the overall story is “unrestrained growth”.

That’s what you expect to see early in the growth of something that will one day be much bigger: first, there’s exponential growth, then as environmental factors start stunting the growth (e.g. the friction that ASCAP, BPI, RIAA, MPAA, and Microsoft among others are very highly motivated to create) the growth slows down to linear, then sub-linear growth, eventually capping out at an “equilibrium” state.

If free culture were an edge phenomenon, only interesting as a loss-leader to attract proprietary sales, for example, you would expect to see this equilibrium be pretty small. However, it’s clear from the numbers that whatever the equilibrium is, we are currently very far from it—exponential growth means that anything that is trying to retard the growth isn’t working, and so there is apparently loads of room to grow.

“Exponential growth” means that over any fixed interval of time, the amount increases by a fixed percentage—you’ve probably heard of this in connection to compound interest. The growth we see here is around 0.6% per day (it’s about 0.53 for all CC works and maybe 0.60 for free works). That probably doesn't sound like much, but it builds up quickly. One way to characterize exponential growth is to specify the “doubling time”—how long do you have to wait for the amount to increase by 100%, or for the phenomenon to double in size?

For 0.6% per day, that number is about 115 days, or a bit under 4 months. Another way of looking at it is that the number is increasing by almost an order of magnitude (more exactly, about 8.9 times) each year. This is the growth rate of free-licensed CC works.

The fact that the free licenses (Public Domain, CC-By, and CC-By-SA) are showing the same growth (in fact, slightly beating the overall CC licensed growth) appears to refute the widely-held belief in the free-software community that the non-commercial and non-derivative variants are stunting the growth of free licenses for cultural works.

Certainly, we can afford to lay off of CC about it: even if they are stunting that growth, we have little to worry about. The free license options are becoming slightly more popular: between early 2005 and the middle of 2006, they went from about 22% to about 29% of the total. Copyleft (By-SA) licensed works are growing just a teensy bit faster than the non-copyleft licensed works, though not enough to really call it significant.

In June 2006, the number of free-CC-licensed works was about 40 Million. If the growth rate observed in this plot keeps up, that number will be about 360 Million by June 2007, and of course, about 3.2 billion by June 2008. After that, I’m a little afraid to predict, but I think you can see that it will be pretty huge.

It’s much more speculative to talk about the relative increase in free-licensed works (PD, By, By-SA), but if the growth curves I've plotted here were to be extended to June 2008, then free works would make up about half of all CC works by then (the total would be about 6.6 billion by then).

Unfortunately, I don’t have comparison figures for the amount of work under proprietary licenses on the web. It’s probably a pretty huge number, so these may well still be small percentages of the total—though it’s already clear that CC licensing is becoming more mainstream, and another couple of orders of magnitude of increase will surely make them much more important to mainstream culture.

Another problem, of course, is that this picture doesn’t distinguish by media. There are some other statistics by media type available, but they’re somewhat limited, and they I’m not able to find time-series data, so it’s not really possible to try to find a trend, though we could naively assume that the media makeup isn’t changing that much.

All in all, it’s a picture of roaring success. Congratulations are clearly due to the Creative Commons!


Copyright ©2007 Terry Hancock / Creative Commons Attribution-ShareAlike License (

Originally published at

You must retain this notice if you reprint this article.

Except where otherwise noted, the illustrations in this piece are by me, and under the same terms.


Verbatim copying and distribution of this entire article are permitted worldwide, without royalty, in any medium, provided this notice is preserved.