A short-term forecasting model for sectoral regional employment

Abstract

Problems associated with employment have become a major issue in regional economics, especially in those countries, such as Spain, where unemployment represents a serious threat to regional stability. In this context, the purpose of our paper is twofold: first, to develop a prediction instrument for regional employment, one which should be flexible and efficient; secondly, to use this instrument in order to make a short term forecast of sectoral regional employment (and the labour force) of a characteristic region of Spain, namely Aragón.The type of instrument used is an econometric time series model specified in a regional context. This means that attention must be given to the spatial structure of the labour markets. One way to introduce this effect may be through the specification of a transfer function which connects the regional market to the national one. This approach is an improvement on the traditional technique of univariate time series. Furthermore, for small labour markets it is important to assess the incidence of outliers, especially for forecasting purposes, because they distort the statistical inference.

Short‐term water dynamics in Chihuahua City, Mexico

“…Following parameter estimation, standard diagnostic checking is carried out to examine what input variable lags should be retained. Under the LTF approach, any remaining systematic movements in the dependent variable are then modeled using a combination of autoregressive and moving average parameters [Liu and Hanssens, 1982;Trívez and Mur, 1999]. Several rounds of diagnostic checking and re-estimation are generally required before a final transfer model specification is selected [Box and Jenkins, 1976;Wei, 1990].…”

“…To carry out the analysis, a linear transfer function (LTF) procedure is employed [Liu and Hanssens, 1982]. The latter has been shown to be useful in regional econometric research due to its emphasis on the roles played by right-hand-side regressors in time series modeling [Trívez and Mur, 1999]. In addition to the standard diagnostics, model reliability is also assessed via a series of out-of-sample simulation exercises [Leamer, 1983;McCloskey and Ziliak, 1996].…”

Maquiladora Employment Dynamics in Nuevo Laredo

“…Several rounds of reestimation may be required before a final model specification is selected. (For a more comprehensive discussion, see Trívez and Mur 1999.) A general functional form for a transfer function ARIMA ( p , d , q ) model specification with two independent regressors ( x and z ) is given by:…”

“…Because it first examines potential correlations between input series and dependent variable, the LTF procedure can handle multiple regressors with relative ease. Trívez and Mur (1999) successfully employ the LTF approach to predict employment patterns in the Aragon region of Spain. The LTF methodology has also been applied to residential energy consumption (Liu and Lin 1991; Tserkezos 1992).…”