Test methodologies for the unitary root

This paper provides a critical survey on unit root testing procedures frequently used in empirical literature. The importance of the hypothesis of linearity of the deterministic component is stressed because its misspecification may result in the false conclusion that the time series has a unit root. Testing strategies elaborated by Dickey, Bell and Miller (1986), Perron (1988) and Dolado, Jenkinson and Sosvilla-Rivero (1990) are compared and their limit is evaluated. An empirical testing strategy is then proposed and applied to the analysis of the nonstationarity exhibited by real GDP in France. We show that it is possible to find some flexible specifications that enable us to reject the u…

Disparités régionales et interactions spatiales dans l'Europe élargie.

A spatial econometric analysis of geographic spillovers and growth for European regions, 1980-1995

The aim of this paper is to consider the geographical dimension of data in the estimation of the convergence of European regions and to emphasize geographic spillovers in regional economic growth phenomena. In a sample of 138 European regions over the 1980-1995 period, we show that the unconditional /? -convergence model is misspecified due to spatially autocorrelated errors. Its estimation by Ordinary Least Squares leads to inefficientestimators and invalid statistical inference. Using spatial econometric methods and a distance based weight matrix we then estimate an alternative specification, which takes into account the spatial autocorrelation detected and leads to reliable statistical i…

Geographic spillover and growth (a spatial econometric analysis for european regions)

The aim o f this paper is to integrate the geographical dimension o f data in the estimation o f the convergence o f European regions and emphasize the importance o f spatial effects in regional economic growth phenomena. In a sample o f 122 European regions over the 1980-1995 period, we find strong evidence o f spatial autocorrelation in the unconditional /^-convergence model using spatial econometric methods with different weight matrices: a simple contiguity matrix and 4 distance-based matrices. Therefore, this standard beta-convergence model exhibit misspecification, its estimation by OLS leads to inefficient estimators and invalid statistical inference. We suggest then a “minimal” spec…

Intra-urban spatial distributions of population and employment : the case of the agglomeration of Dijon, 1999

The aim of this paper is to analyze the intra-urban spatial distributions of population and employment in the agglomeration of Dijon (regional capital of Burgundy, France). We study whether this agglomeration has followed the general tendency of job decentralization observed in most urban areas or whether it is still characterized by a monocentric pattern. In that purpose, we use a sample of 136 observations at the communal and at the IRIS (infra-urban statistical area) levels with 1999 census data and the employment database SIRENE (INSEE). First; we study the spatial pattern of total employment and employment density using exploratory spatial data analysis. Apart from the CBD, few IRIS ar…

Une stratégie de test de la racine unitaire

This paper compares three approaches of the unit root testing strategy issue : i.e. the DICKEY, BELL and MILLER (1986) approach, elaborated for prevision purposes, which advocates the estimation of a model without trend at the outset, in opposition to the PERRON (1988) and HENIN and JOBERT (1990) testing strategies. The latter is undoubtedly operational in empirical works but it doesn’t encompass joint tests explicitly. We then suggest an unifiedgeneral framework embracing the last two appoaches and giving a practical scheme which is as complete and operational as possible. The behaviour of these strategies is assesed in an empirical study of the seasonaly adjusted real GDP in France and an…

Non-stationarity tests and nonlinear trends

This paper stresses the importance of the hypothesis of linearity of the deterministic component imposed by unit root testing procedures most frequently used in empirical literature. We suggest an empirical testing strategy which reduces the risk of reaching false conclusions due to the misspecification of that component and we apply it to the analysis of the nonstationarity exhibited by real GNP in France. We show that it is possible to find someflexible specifications which enable us to reject the unit root null hypothesis otherwise strongly supported in empirical literature. These specifications might be considered as approximations of the true process generating real GNP and might be us…