Changes

*They are especially ineffective at protecting process innovations

*'''Legal requirements''' for upholding their validity or proving their infringement '''are high'''

'''Trade secrets''' may be a viable alternative if the product does not reveal the technology (or it can not be reverse engineered). This is likely to be true when knowledge is '''tacit''', rather than '''codified'''. Tacit knowledge is by definition hard to articulate, and can only be transmitted by demonstration. Simplistically, an appropriability regime might be regarded as '''strong''' or '''weak''', with regard to good the protection is.

===Dominant Design Paradigm===

*'''Cospecialized''' assets with a '''unilateral dependence''': Either the innovation needs to be specialized to work with the asset, or vice versa

*'''Cospecialized''' assets with '''bilateral dependence''': Both the innovation and the asset need to be specialized to work with each other

Cospecialized assets need '''relational specific investments''' from one or both side (c.f. Williams). In the context of '''incomplete contracts''' (Teece doesn't mention this explicitly), there will be the possibility of '''hold up''', making contracting (without fiat) '''costly and difficult'''.

*'''Contracting'''

*'''Integrating'''

Under contracting the innovator will not have to make '''up front payments''' to build or buy the asset, which '''reduces risk''' and capital requirements. The contract can be for a '''strategic partnership''' rather than a '''simple buy-sell agreement''', and in the former case can bring credibility to the innovator. But, in the paradigmatic phase, this has two problems: