--Summary by James Bessen

Introduction
This paper analyzes the results of a broad survey of participants in Free/Open Source software (F/OSS) projects. Several of the findings challenge common assumptions about programmers’ motivation.

The authors identified a broadly representative sample of programmers from SourceForge, the largest community listing of F/OSS projects. They limited their survey to projects that had at least reached the Alpha stage of development, where the initial code is available, and conducted an additional survey of more mature projects. In all, they received 684 responses from 287 distinct projects.

The study aims to understand what motivates F/OSS development by focusing on the personal motivations of software developers. The survey asked developers to rate the importance of various motives, and the authors then classified these responses into intrinsic and extrinsic motivations. Intrinsic motivations include personal enjoyment, such as the fun and challenge of working on a creative project, as well as a sense of community obligation. Some of the extrinsic motivations involved immediate benefits, including pay and the desire to solve particular problems of specific value (“scratch an itch”). Others provide delayed benefits, including enhanced career prospects from participation and the acquisition of better software development skills.

The authors analyze the survey returns using tabular analysis, factor analysis and regressions.

Results
The authors make several surprising and strong findings:

1. F/OSS developers have a wide variety of motivations.

2. Intrinsic motivations are reported to be more important than extrinsic motivations.
Much academic work, particularly in economics, has emphasized the importance of extrinsic motivations, like career concerns. Surprisingly, developers do not ascribe much importance to these motivations, but instead emphasize personal enjoyment, especially in terms of the creativity involved in a project.

3. Firms are responsible for about half of the development effort.
Much of the analysis to date has assumed that F/OSS is primarily driven by volunteers, with firms making contributions only when they have specific needs—for example, drivers for complementary hardware products. But 38 percent of the respondents do their F/OSS development at work, with the awareness of their supervisors. (Another 17 percent make their contributions at work without the supervisors knowing.) Moreover, since these people tend to put in about 50 percent more hours than other F/OSS developers, they contribute about half of the total effort. In other words, firms account for about half of the development effort, in contrast to common perceptions.

4. Intrinsic motivations matter for paid developers, too.
The first two findings might seem contradictory, but they are not. Performing a regression analysis, the authors find that the level of effort is affected strongly by the creativity of the project even among paid developers. This finding is a nice illustration of the old “principal-agent” problem—eliciting effort from one’s employees is more complex than simply offering them a pay check. And so the motivation of volunteers ends up being rather similar to the motivation of paid developers. The authors further note that perceptions of what is “creative” are highly personal, and projects that one programmer finds fascinating may seem like drudgery to another.

This study is important both because it captures the broad scope of F/OSS activity, and because its findings challenge common assumptions. Clearly, it raises as many questions as it answers. For example, future research will have to explore what motivates firms to support F/OSS, how firms manage F/OSS projects and how different types of projects affect workplace motivation