Seminario 16/01: R. Valero (U. Alicante) – Synthetic Control Method versus Standard Statistic Techniques a Comparison for Labor Market Reforms
- Ponente: Rafael Valero. Universidad de Alicante
- Fecha: 22/Ene/2016 - 12:30 horas
- Lugar: Seminario del Departamento de Métodos Cuantitativos para la Economía y Empresa, UMU. Retransmisión en directo pendiente confirmar.
The aim of this paper is to create counterfactual series in a reliable way and illustrate its use. The present literature is dominated by the Synthetic Control Method (SCM), but there are cases where it cannot perform and it was not created considering the trade-off between bias and variance, hence the accuracy of results could be improved. The first contribution of this paper: I have modi ed the current method to alleviate its weaknesses. The second contribution is that I propose standard statistical alternatives with the main focus on LASSO with Cross-Validations; these are very easy to implement and in-built in most software packages. The third contribution is to apply the previous methods to two Labor Market Reforms: a) reduction of the work week load, I study Portugal (1996); and b) reduction of ring cost, I study Spain (2010). The results are that there is a reduction of unemployment in Portugal. For the Spanish case the law showed an increase in unemployment rate.