Can one understand Stack Exchange sites if looking at them as markets?

Question

Often, when I explain friends, who are economists, about the principles of a Stack Exchange, I discover that they cannot get such an exchange to fit into their framework of a market, and therefore do not believe that Stack Exchanges are viable in the long term. The definition I would like to use for a market is that it is a place where one can exchange something of value (in this case, knowledge) for the benefit of both parties. I understand that, if for example the reputation points are considered the currency in this market, then it is a currency with most peculiar properties.

Answer 1

I hope that none of the economists I work with would reach a conclusion similar to that reached by your economist-friends. To deny that anything that doesn't fit well within a traditional market framework won't work would deny that things like open source software and Wikipedia exist.

Indeed, I think the economics of open source software offers some insights as the OSS and SE paradigms are fairly similar. A gut reaction of many economists is by probably similar to the view espoused in the now famous The Cathedral and the Bazzar (Ramond, 2000). Here we find (after Lerner and Tirole, but they get the citation wrong):

The 'utility function' Linux hackers is maximizing is not classically economic, but is the intangible of their own ego satisfaction and reputation among other hackers ...

If this is the case, then stackexchange has setup an very explicit system for tapping into this drive for "ego satisfaction" -- the reputation system.

Lerner and Tirole's (2002) "Some Simple Economics of Open Source" is still a great place to start thinking about open source and similar movements. Many economists will hold that people are indeed rationally maximizing when they participate in things like open source software and stackexchange, even if the benefits are not quite obvious. L&T state (pp. 212-213):

A programmer participates in a project, whether commercial or open source, only if she derives a net benefit (broadly defined) from engaging in the activity...

The costs of participating in such projects are fairly obvious (namely opportunity cost), but T&R also note the (less obvious) benefits. These are grouped as immediate and long-run benefits and, of the immediate benefits the most relevant here is probably that "A 'cool' open source project may be more fun than a routine task." (That is, it is fun, and it should be thought of as leisure activity). In the long run, benefits derive from increased reputation (and hence cooperation from the community).

But, SE is a bit different from an open-source software project. Here, it might be obvious why people ask questions, but it's not quite so obvious why people answer them. Lakhani and von Hippel (2003) provide some guidance based on surveys of people who answer questions about open source software (specifically Apache). They find that those answering questions are motivated to scan the questions "primarily in order to learn, rather than to answer questions." (p. 935). So, without this learning motivation, many information providers would never even come across questions they can answer--this is likely to be a motivation here too. Once answerable questions are found, the motivations for actually answering include the (now) familiar entertainment and reputation-building incentives, but a common thread is also an inclination towards "general" reciprocation. It seems that contributors desire to help the community that helped them, even if they cannot help the individual who helped them. (General reciprocity is not a new concept in economics)

Or, maybe we are just zealots: The Quality of Open Source Production: Zealots and Good Samaritans in the Case of Wikipedia

Answer 2

There are lots of behaviours that are (initially) difficult to shoehorn into a framework of a market with supply, demand and a medium of exchange.

Nevertheless, there are economic insights to be gained by looking at patterns of motivation and (perceived) rewards.

The system of apparently-random, frequent, variable rewards is very addictive. c.f. fruit (slot) machines; and behavioural-economics studies of addiction

The relationships between motivation, intrinsic rewards and extrinsic rewards has been discussed before - see meta.stackoverflow.com passim.

Other motivations for contributors, that could fit into a classical- or behavioural-economics framework:

• altruism;
• paying-forward by contributing valuable content now, in the expectation of having questions that will need answering later;
• perceived opportunities for career advancement;
• learning by trying to solve problems, and having one's solutions challenged;
• learning by investing time and effort, rather than money;
• a place to keep one's mind honed, while it is otherwise underemployed.

Answer 3

In a financial market, trade is based on mutual exchange of capital, whether in products/services or cash. However, Stackexchange is closer to an academic market. Just as professional academics are valued on their number of published papers and the citations of those papers for their credibility and prestige, so in Q&A markets.

Here's Robert Scoble on using Stackoverflow for recruitment.

For working, non-academic, professionals who may be very good at programming, finance or economics (amongst the wide range of SE sites) being able to achieve recognition may assist with professional advancement. Obviously, asking questions assists in passing through the steep learning curve in these professions a little faster...

Answer 4

One of my favorite papers on this topic studies Stackoverflow and the effect of badges on inducing participation. What differentiates this study relative to many others is that it uses quasi-experiments to try to get at causality. It's not a perfect study (what is?), but it definitely represents a good stab at the question.

Causal Discovery in Social Media Using Quasi-Experimental Designs http://www.cs.umass.edu/~hoktay/pub/soma2010.pdf