In 2006 was approved, in Brazil, a new regulatory framework, named General Law of Micro and Small Enterprises (MSEs), regulating and expanding special conditions for MSEs - that representing more than 90% of companies in Brazil. The Law introduces differential treatment to MSEs for starting a business, reducing legal requirements, and providing differential treatment for government purchases, among others. In São Paulo state, the Law was promulgated in late 2007. About 340 municipalities (out of 648) approved local laws in the same way of National one. Among these, in 91 municipalities already there evidence that these points are actually implemented. The objective of this study is to estimate the effect of regulation of the MSEs' General Law in municipalities of the state of São Paulo over economic indicators. We use a treatment effect analysis, and we distinguish the municipalities that received treatment (implemented the local laws) who did not (other municipalities). The municipalities, however, do not have the same probability of belonging to one group or another. Observables variables can distinguish the two groups, such as the importance of industry and specialized services, activities of trade associations, formalization rate of hand labor, number of micro and small enterprises before implementation, and etc. Specifically, factors such as location and proximity to other cities that implemented the Act may also influence. In a sense, the spatial distribution of the implementation of Law over time can also be considered as an important factor for future implementations. This fact introduces a selection bias when comparing sets of municipalities whose possibilities were different and that therefore the best answer is different. We use a spatial propensity score matching to control the potential effect of neighbor municipalities through the probability of selection. Thus, we select those municipalities that have not implemented the law with similar observables characteristics to cities that implemented, to constitute a control group for purposes of evaluating the performance of treated municipalities. As the effect of the treatment can be given in time, can also, depending on the available database, subdivide the group of municipalities treated over time, making sure the effect is greater in those who longest time implemented the Law.
