States will follow the imitation effect to establish equal welfare distribution between classes – empirically proven by spanish regions
Ayala et al. ‘21 (Researcher at Georgia State University) ,"Georgia State University, https://icepp.gsu.edu/files/2021/03/21-07-Interregional-Mimicking-and-Decentralized-Welfare.pdf
The literature on welfare decentralization has traditionally stressed the potential positive effects of fiscal devolution both in terms of efficiency and coverage of the programs. Regional governments are in a better position to understand both social preferences and needs of poor households and generally they can implement these programs more effectively. However, the expectation that a decentralized provision of welfare is supposed to foster citizens´ wellbeing is challenged by problems of coordination and financing, which at the end may produce a mosaic of highly varied programs—with a striking disparity of protection levels. In addition, competition among jurisdictions does not always yield the result of positive innovation. Ignoring these constraints can result in a generally regressive nationwide distribution of benefits, with the richest jurisdictions paying much higher benefits than the less wealthy ones. These limitations, common to any decentralized welfare system, can be fostered in 28 models of “radical fiscal federalism”, where federal coordination and/or funding do not exist. This was the case of the Spanish safety net design, where these programs were entirely created and regulated by the regional governments. In this paper, we use panel data for Spanish regions with the aim of answering one essential question: Does mimicking among ACs partly explain the level of regional welfare benefits in Spain? While the answer to this question is not a priori obvious, our empirical results corroborate the presence of significant interregional interactions. We find strong evidence of a mimicking behavior: ACs observe what their neighboring governments are doing and then decide their own basic benefit levels. Therefore, our results confirm the conventional wisdom on the territorial interdependency of policy decisions with respect to welfare benefit levels, even within a highly fiscally decentralized framework—where there is no participation of the central government in the design, regulation, or financing of the system. Therefore, mimicking provides a vehicle that can save fully decentralized systems of welfare benefits from the expected race-to-the-bottom. Our results suggest that the divergent trend in welfare benefits triggered by regional differences in fiscal capacity is partially compensated by regional interdependence. Our results are obtained utilizing the spatial lag approach and confirmed when the proposed relationships are analyzed using the spatial error approach. Given the probable inertia of benefit levels, the results obtained under static approaches could be somewhat biased. However, the use of a dynamic model also suggests that, while benefit levels in each region largely depend on regional resources, the mimicking hypothesis is still validated: regional public agents observe what their peers are doing and act accordingly. In short, in this paper we contribute to the current literature by providing strong supporting evidence for the role played by imitation in the determination of decentralized 29 welfare benefits in contexts where these policies are completely decentralized. Therefore, we confirm the results anticipated by other studies that governments pay attention to what their neighbors do, but in our case with the novelty of confirming it even when there is no system of coordination or co-financing from the central government whatsoever.