Essays on patent economics

In Chapter 1, new examination process indicators are used to revisit the literature on the national treatment principle. Based on lower grant rates, several studies present evidence suggesting that patent offices are biased in favor of their national applicants. If true, this “national bias” would be proof of disrespect of several international patent-related treaties. Econometric results from Chapter 1 are aligned with the literature: domestic applicants have a lower probability to be granted a patent than their foreign counterparts, at the EPO and the JPO. However, no evidence of national bias is found throughout the work of the patent offices. This chapter suggests that grant rates analyses may induce misleading conclusions on national bias, mainly because they are affected by economic forces, which impact the willingness and/or ability of the applicant to maintain its patent active. Indeed, as illustrated by de Rassenfosse and Hosseini (2020), accessing foreign markets is always more complicated and costly, be it for systemic, languages, or market differences. The chapter concludes that further research is necessary to measure the impact of economic forces on applicant behavior and patent examination outcome.In the second chapter, an empirical methodology is introduced to test whether and to what extent search and examination processes at patent offices are substitutes. Two groups of applications are compared: those for which patent offices have access to the search for prior art of other offices, and those who start from scratch. When patent offices have access prior work, they work “less” (they search into less technology classes, generate less backward citations, and send fewer communications) and differently (they rely more on international citations and provide more information early in the process). This paints the picture of a much more connected international patent examination network than generally believed, where patent offices trust the work carried out by their colleagues. Further substitution could still be achieved, as around 55% of technology classes searched and up to 70% of backward citations are duplicates -voluntarily or not - of prior examination work. Adding the reduction in the amount of work carried out and the remaining overlap together, the estimates show a degree of substitution of about 37% between the three patent offices. However, to fully understand how synergies can occur, it would be useful to know if the extra work carried out by the patent office is essential to the final decision.The findings from Chapter 2 are particularly relevant to advance the discussion raised at the end of Chapter 1. In fact, if patent examination processes are substitutes, the price paid by the applicant has a direct impact on the process he will choose to submit (and then maintain) his invention to (especially in PCT applications where the first patentability assessment is centralized). The significant substitution level estimated in Chapter 2 confirms that cost can be a significant factor in the applicant’s commitment to the examination process. This supports the necessity to take economic forces into account when comparing examination process outcomes. Chapter 3 answers this by measuring the extent of the economic factor in the three patent offices. When an application is not granted by a patent office, three types of failure are identified: economic withdrawals, induced withdrawals, and refusals (explicit or implicit). The classification is based on how strongly the office reports contest the patentability of the invention, the timing of the exit, and the reactiveness of the applicant. The empirical analysis reveals that a third of the patent applications that failed to get a grant are likely caused by economic forces. Econometric models further estimate that the economic factor is significantly more present at the EPO, and when the failed outcome is an exception within the patent family (that is, the invention is granted in the other patent offices). The prevalence of economic withdrawals is so great that it challenges the adequacy of grant rates to discuss office stringency and encourages the inclusion of a wider array of applicant-based indicators. The main question that the analysis in Chapter 3 fails to answer is the reason why the EPO presents such a higher proportion of economic withdrawals. One assumption that needs to be explored is whether or not it can simply be attributed to a longer examination process. In fact, if the annual economic withdrawal rate was similar across patent offices, the longer duration of the EPO process could lead to a higher proportion of economic failures. However, if this hypothesis is rejected, a residual explanation is that the cost-benefits ratio might be higher than at the American or Japanese offices.In Chapter 4, a time dimension is included to explore the evolution of economic forces over time, across offices, and across applicant characteristics. Based on Kaplan-Meier survival functions and Cox proportional hazard ratios, EPO applications are estimated to withdraw earlier and in greater proportion, especially compared to USPTO ones. This effect is still present after the patents are granted, and is robust to invention- and applicant-related characteristics. This paper argues that the lower grant rates observed at the EPO might be explained by a lower cost-benefits ratio that impacts the applicants’ commitment to the examination process.To sum up, the main contribution of this thesis is that it explores the patent examination process in detail, which helps distinguishing the contribution of the patent office from the role of the applicant and provides a better understanding of the drivers behind patent examination outcomes. Achieving this goal was only feasible because of the creation of a unique new database.