Template-Type: ReDIF-Article 1.0
Author-Name: Oscar Luis Alonso Cienfuegos
Author-Name: Ana Isabel Otero Sánchez
Title:  UN ANÁLISIS ESPACIO-TEMPORAL DE LA POLÍTICA AGRARIA COMÚN EN ASTURIAS PARA EL PERIODO DE PROGRAMACIÓN 2007-2013 Y SU INCIDENCIA EN EL EMPLEO.
Abstract: Resumen:Se pretende contrastar con este trabajo los resultados obtenidos, en términos de empleo, por los fondos públicos de la Política Agraria Común destinados a dinamizar las zonas rurales de Asturias durante el último periodo de programación finalizado. Se plantea la consideración de los ámbitos espacial y temporal, utilizando modelos espaciales de datos de panel, para estudiar el comportamiento y las posibles relaciones existentes entre fondos públicos y empleo, durante el periodo de programación 2007-2013. Nos centraremos especialmente en el modelo Durbin espacial.
Abstract:It is intended to contrast with this work the results obtained, in terms of employment, by the public funds of the Common Agricultural Policy aimed at boosting the rural areas of Asturias during the last programming period completed. It arises the consideration of spatial and time ambits, using spatial panel data models, to study the behaviour and possible existing relationships between public funds and employment, during programmation period 2007-2013. We will focus especially in spatial Durbin model.
METHODOLOGY
We will use the spatial econometrics approach to try to determine the incidence of public funds of the CAP in Asturias for the 2007-2009 programming period, by estimating spatial models of panel data that quantify the possible positive relationship between public investment carried out and employment.
Although we do not consider it convenient to estimate models without a theoretical basis, we do not consider it convenient to simply extend or adapt, without more, to spatial context, existing theoretical models, as extensions of them, without considering all the empirical problems that they may present. We consider the development of spatial econometrics, taking advantage of the solutions proposed by recent advances and adapting the theoretical models to them, and not vice versa, to give a direct solution to the problems that may arise. Some authors review the situation of spatial econometrics and the important changes experienced in recent years that have improved the estimation techniques and methodology available, as well as the tools and software packages developed. These recent advances have been used to directly raise models that explain the impact of public funds framed in the CAP for the case of Asturias in the last scheduled execution period. The theoretical foundation arises from the premise that justifies the implementation of these economic policy measures, and this is that investment or public spending in the primary sector and rural areas creates employment. This would suppose a positive and significant effect of said explanatory variables in the variable explained, employment. Together with these, we will take into account other variables that the literature has considered relevant when it comes to modeling these aspects.
We will use for our study a model of panel space time, to try to fix the relations, as well as the diffusion effects and spillovers in both dimensions (in time and in space). We will also use a decision process in the selection of models from the general to the particular, that is to say we will start from the general specification and then impose restrictions on the parameters, either by theoretical criteria or by estimation criteria, to consider the best model of the possible ones within a model nesting structure, in which we pay special attention to the Durbin model, by the possible advantages that present the estimators obtained in his case.
Mohl and Hagen (2011) consider that from a theoretical point of view the payments of the European structural funds can affect employment through the demand for work by increasing the public and private capital in the region. Another channel of influence would be through the increase of technological progress.
We will also include in our case, together with the fundamental variable of public funds, a Xit vector of socioeconomic variables that have been contrasted in the literature as relevant and with clear causality for determining the level of employment.
We will initially propose a spatial model of panel data similar to that described in (1) and (2), to consider the temporal and spatial dimensions
Spatial model of panel data
y_t=ρWy_t+x_t β+Wx_t γ+η+ξ_t τ+ε_t
ε_t=λWε_t+u_t      
Where η: vector of unobservable individual effects;
ξ: non-observable temporal effect corresponding to the period of time t;
τ: vector of dimension ones (n × 1).
The joint control of the unobservable effects could cause us problems, therefore, we will only consider unobservable individual effects, without including the temporary effects, to avoid a significant loss of degrees of freedom, as well as the possible existence of multicollinearity, which could distort the results of the estimates.
For the selection of the spatial weights matrices we will compare the results of the models with different types, specifically with matrices of contiguity relations with 4, 5 or 6 neighbors based on the inverse or binary distance, and standardized by rows or without standardizing.
RESULTS
The initial general model to be estimated, in a process of estimating the general to the specific, has the following functional form:
〖emp〗_t=ρ_ W〖emp〗_t+〖ipac〗_t β_1+〖pob〗_t β_2+〖empagri〗_t β_3+〖paro〗_t β_4+W〖ipac〗_t γ_01+W〖pob〗_t γ_02+W〖empagri〗_t γ_03+W〖paro〗_t γ_04+η+ε_t        
ε_t=λWε_t+u_t
1. Public funds: IPAC is the sum of the amount of public funds allocated to rural development (second pillar of the CAP), plus the amount of public funds destined for the first pillar of the CAP, in Asturias for the years 2007 to 2013 by councils (municipalities).
2. Population: POB is the number of inhabitants according to the census by councils (municipalities) for Asturias in the years 2007 to 2013.
3. Employment: EMP is the number of jobs by councils (municipalities) for Asturias in the years 2007 to 2013. EMPAGRI is the number of jobs in the primary sector by councils (municipalities) for Asturias in the years 2007 to 2013.
4. Unemployment: PARO is the number of unemployed by councils (municipalities) for Asturias in the years 2007 to 2013
We use the Panel Data Toolbox v2.0 package developed by Álvarez et al. (2016) for Matlab. The estimation method is that of instrumental variables, if only the spatial delay of the endogenous variable is included, and that of generalized moments, if space delay is included in the error structure.
The different models are proposed, Durbin, Durbin error, lag and spatial error. Starting from the nested general model, we arrive at the conclusion that the most appropriate is the spatial Durbin model, with random individual effects, with spatial lag in the variable "pob", and random effects, in view of the results of the Hausman test, that do not allow to reject the null hypothesis of non-existence of systematic differences in the estimators of the model of fixed effects and those of the random effects model.
The results of other options are also exposed, to compare them, and also to compare the possible behavior of the variable related to public investment in the Common Agricultural Policy in Asturias, under other possible specifications. There is a clear lack of significance of the variable "ipac", which does not allow its repercussion to be considered relevant, at least in favorable terms for employment dynamics. The coefficient of "ipac" is not significant in any of the proposed models, except in the model of delay and spatial error, estimated by random effects, but it is the only case in which the Hausman test clearly indicates rejecting the null hypothesis and choosing the fixed-effect model versus the random-effects model.
CONCLUSIONS
The analysis carried out, based on the panel data methodology, to consider both dimensions, space and time, allowed us to study the behavior or possible impact of public funds allocated to the CAP in Asturias on one of the most important variables. relevant and among its priority objectives, as a key to the maintenance of rural territories, employment. The results have been negative, confirming the lack of significance of the variable "ipac". An estimation process has been carried out from the general to the particular and the spatial Durbin model has shown the best results. The estimation method used has been that of instrumental variables. In the cases in which space lag was included in the error term, generalized moments have been used. The rest of the variables considered are relevant, as well as the proposed spatial relationships, confirming the strong relationship between employment, unemployment and population. The lack of significance of the investment made is something also expected if we analyze the current situation of rural territories and the generalized negative behavior of the evolution of the population and employment variables within them. This trend has been the one that has driven economic policy measures to try to reverse the situation, but it does not seem that, at the moment, at least according to the estimations of this study, one can speak of obtaining good results in terms of employment.
Classification-JEL: R1
Keywords: Econometría espacial, Desarrollo rural, Datos de panel, Modelo Durbin Espacial., Spatial econometrics, rural development, Panel Data, Spatial Durbin Model.
Pages: 39-65
Volume: 3
Year: 2019
File-URL: http://www.revistaestudiosregionales.com/documentos/articulos/pdf-articulo-2578.pdf
File-Format: Application/pdf
Handle: RePEc:rer:articu:v:3:y:2019:p:39-65