BOBTCHEFF Catherine

In the first chapter of the thesis, I study the impact of rebuttal incentives on publication quality. One lesson to be learned from the analysis conducted is that the impact of the rebuttal reward on publication quality depends on the difference in speed between pioneering research and rebuttal activities. If the former is greater than the latter, publication quality decreases with the reward for rebuttal. Otherwise, the opposite is true. Compared to this basic model, publication quality is always lower when publication quality is uncertain. This last result suggests that transparency improves research quality. In the remaining two chapters of the thesis, I examine the trade-offs of scientists accessing citizen science (CS) or traditional science (TS) to conduct their research. By using CS, scientists involve the help of citizens, which is expensive and makes the research process faster. However, the quality of the results obtained is diminished. The analysis in the second chapter shows that if a scientific project is socially desirable, a scientist without a competitor chooses CS. Otherwise, he chooses ST. If the latter has a rival, the two scientists compete only if they both choose the SC. An interesting result emerges if scientific projects are moderately socially desirable: scientists randomly choose either SC or ST. The public policy implications depend on the scientists' choice. If one chooses CS and the other ST, there is no loss of efficiency. This result suggests a laissez-faire policy. If both choose ST, the research process is slowed down compared to the case where the scientist has no competitor. Therefore, the government should encourage CS through grants. In the third chapter, the scientists fight to establish publication priority for an innovative idea. Unlike the second chapter, the two scientists compete if they both choose ST. CS is an option they can exercise at the time they discover the idea. One lesson to be learned from the analysis conducted is that the scientists' choice between CS and ST does not depend on when they discover the idea. If SC has a low cost, they both choose SC. If not, they both choose ST. Compared to a research environment where it is not possible to choose CS, these results have two implications. First, scientists do not have time to adapt to this new research opportunity. Second, it increases the fear of being pre-empted. Therefore, in terms of public policy, the government must encourage the use of CS if it wishes to increase the speed of publication.