Transparency as design choice of open data contests

Open data contests have become popular virtual events that motivate civic hackers to design high performing software applications that are useful and useable for citizens. However, such contests stir up controversy among scholars and practitioners about the role of transparency, or more specifically, the unrestricted access and observability of the applications submitted throughout the contest. In one view, transparency may reduce performance because it causes excessive replication, whereas another view argues that transparency can encourage novel forms of reuse, namely recombination. This article proposes a new perspective towards transparency as a design choice in open data contest architectures. We introduce a 2‐dimensional view towards transparency, defined as observability of information about each submitted (a) solution (how it works) and its (b) performance (how high it scores). We design a sociotechnical contest architecture that jointly affords both transparency dimensions, and evaluate it in the field during a 21‐day contest involving 28 participants. The results suggest that the joint instantiation of both transparency dimensions increases performance by triggering different kinds of recombination. Findings advance literature on sociotechnical architectures for civic design. Furthermore, they guide practitioners in implementing open data contests and balancing the tension between individual versus collective benefits.