When deciding whether to share information, firms consider their private welfare. Discrepancies between social and private welfare may lead firms excessively to share information to anti-competitive ends - in facilitating of cartels and other harmful horizontal practices - a problem both antitrust scholarship and case law have paid much attention to. On the other hand, legal scholars have paid far less attention to the opposite type of inefficiency in information sharing among competitors - namely, the problem of sub-optimal information sharing. This phenomenon can generate significant social costs and is of special importance in network industries because the maintenance of compatibility, a key to producing positive network effects, typically requires information sharing. Understanding the hitherto neglected impact of sub-optimal information sharing is important not only for many areas of antitrust law, but also for developing effective policies towards network industries and critical infrastructures more generally, as well as for improving those procedural rules that concern information exchange among litigating parties. This paper therefore advances the legal analysis of impediments to efficient information sharing in a number of significant ways: First, it shows that the strategic behavior of competitors may erect an economic barrier to information sharing that has not been previously addressed in the literature - the fear of degradation. This form of strategic behavior involves the strategic refusal to share information when the refusal inflicts a greater harm on one's rivals than on oneself, and thus generates a competitive advantage. Second, the paper reveals a hitherto unrecognized set of behavioral impediments to information sharing, wherein rivalry norms and managers' risk attitudes bias competitors' judgments of the prospects of information sharing and the status-quo bias and ambiguity aversion lead these decision makers to avoid such arrangements. Third, it integrates these economic and behavioral insights with the findings of the extant literature to create a new framework for predicting when private information sharing will be suboptimal. Finally, we suggest how the alignment of private information sharing with social optimality may be promoted, based on the framework developed here.