Most patent systems allow applicants to defer patent examination by some time. Deferred examination was introduced in the 1960s, first at the Dutch patent office and subsequently in many other countries, as a response to mounting backlogs of unexamined patent applications. Some applicants allow the examination option to lapse and never request examination once they learn about the value of their invention. Examination loads are reduced substantially in these systems, albeit at the cost of having a large number of pending patent applications. Economic models of patent examination and renewal have largely ignored this important feature to date. We construct a model of patent application, examination and renewal in which applicants have control over the timing of examination and study the tradeoffs that applicants face. Using data from the Canadian patent office and a simulated GMM estimator, we obtain estimates for parameter values of the value distributions and of the learning process. We use our estimates to assess the value of Canadian patents as well as applications. We find that a considerable part of the value is realized before a patent is even granted. In addition, we simulate the counterfactual impact of changes in the deferment period. The estimates we obtain for the value of one additional year of deferment are relatively high and may explain why some applicants embark on delay tactics (such as continuations or divisionals) in patent systems without a statutory deferment option.
patent patent value value of patent applications patent examination deferred patent examination