Studies have shown that if you add in even a small amount of extrinsic motivation (i.e. pay people) it dramatically affects their perception of intrinsic reward.
These studies are typically set up such that there's no choice; you have kids (usually) do some activity like drawing for either no reward or extrinsic reward, and then see how many continue once you remove the extrinsic reward.
There are a lot of possible explanations for this; the one I like the most is based on self-perception theory, which would predict that once you've accepted the reward, you attribute your motivation to the reward.
There are some differences between those studies and this scenario:
1. Open-source developers have a far wider range of options, and must actively choose to work on a project with financial reward.
2. Open-source developers who have already contributed some amount for free should be less influenced than new developers -- they already have had altruistic/intrinsically-motivated self perceptions.
Developers that are attracted to OWS projects because of this reward probably won't stay if the reward is removed, but I'm not sure, and don't predict, that OWS projects will attract fewer developers overall because of this.
I think that intrinsic and extrinsic motivators can definitely co-exist, especially for domains where there are differences in the tasks -- for example, offering these bounties for "less fun" stuff (or dynamically offering more for parts of the repo that have been touched further into the past).
These studies are typically set up such that there's no choice; you have kids (usually)
There's a lot more literature behind this than just studies with kids, though I'm a total novice here. Here's[1] a painfully detailed meta-analysis I found.
Developers that are attracted to OWS projects because of this reward probably won't stay if the reward is removed, but I'm not sure, and don't predict, that OWS projects will attract fewer developers overall because of this.
Reducing this down to just "more or fewer contributors" is the kind of over-simplifying that I worry about here. The software industry is not historically people-savvy, and open source is an entirely social phenomenon. It's easy to mess things up.
I'm less worried about the number of people changing than I am their relationship with the projects they work on changing. Right now, if I send someone a patch, I feel good because I know I'm doing a purely good thing. If I send a patch to a project that pays for patches, that's not bad, but it means now I feel more like I'm freelancing or some other more complication jumble of emotions.
Likewise, if someone sends me a patch, do I get excited that they want to be a part of my project? Have I met a kindred spirit? Or are they just gunning for a chunk of BTC in their off hours?
I'm not saying introducing money is a bad thing here, but like all other avenues of social activity, involving even a small amount of money radically changes the nature of it. Don't believe me? Try tipping your parents after they have you over for dinner. Maybe leave a fifty on your significant other's bedside table and see how that goes. :)
Comments
These studies are typically set up such that there's no choice; you have kids (usually) do some activity like drawing for either no reward or extrinsic reward, and then see how many continue once you remove the extrinsic reward.
There are a lot of possible explanations for this; the one I like the most is based on self-perception theory, which would predict that once you've accepted the reward, you attribute your motivation to the reward.
There are some differences between those studies and this scenario:
1. Open-source developers have a far wider range of options, and must actively choose to work on a project with financial reward.
2. Open-source developers who have already contributed some amount for free should be less influenced than new developers -- they already have had altruistic/intrinsically-motivated self perceptions.
Developers that are attracted to OWS projects because of this reward probably won't stay if the reward is removed, but I'm not sure, and don't predict, that OWS projects will attract fewer developers overall because of this.
I think that intrinsic and extrinsic motivators can definitely co-exist, especially for domains where there are differences in the tasks -- for example, offering these bounties for "less fun" stuff (or dynamically offering more for parts of the repo that have been touched further into the past).
There's a lot more literature behind this than just studies with kids, though I'm a total novice here. Here's[1] a painfully detailed meta-analysis I found.
Reducing this down to just "more or fewer contributors" is the kind of over-simplifying that I worry about here. The software industry is not historically people-savvy, and open source is an entirely social phenomenon. It's easy to mess things up.
I'm less worried about the number of people changing than I am their relationship with the projects they work on changing. Right now, if I send someone a patch, I feel good because I know I'm doing a purely good thing. If I send a patch to a project that pays for patches, that's not bad, but it means now I feel more like I'm freelancing or some other more complication jumble of emotions.
Likewise, if someone sends me a patch, do I get excited that they want to be a part of my project? Have I met a kindred spirit? Or are they just gunning for a chunk of BTC in their off hours?
I'm not saying introducing money is a bad thing here, but like all other avenues of social activity, involving even a small amount of money radically changes the nature of it. Don't believe me? Try tipping your parents after they have you over for dinner. Maybe leave a fifty on your significant other's bedside table and see how that goes. :)
[1]: http://www.rug.nl/gmw/psychology/research/onderzoek_summersc...