IDEAS home Printed from
MyIDEAS: Login to save this article or follow this journal

New Neural Network Methods for Forecasting Regional Employment: an Analysis of German Labour Markets

  • Roberto Patuelli
  • Aura Reggiani
  • Peter Nijkamp
  • Uwe Blien

Abstract In this paper, a set of neural network (NN) models is developed to compute short-term forecasts of regional employment patterns in Germany. Neural networks are modern statistical tools based on learning algorithms that are able to process large amounts of data. Neural networks are enjoying increasing interest in several fields because of their effectiveness in handling complex data sets when the functional relationship between dependent and independent variables is not specified explicitly. The present paper compares two NN methodologies. First, it uses NNs to forecast regional employment in both the former West and East Germany. Each model implemented computes single estimates of employment growth rates for each German district, with a 2-year forecasting range. Next, additional forecasts are computed, by combining the NN methodology with shift-share analysis (SSA). Since SSA aims to identify variations observed among the labour districts, its results are used as further explanatory variables in the NN models. The data set used in our experiments consists of a panel of 439 German (NUTS 3) districts. Because of differences in the size and time horizons of the data, the forecasts for West and East Germany are computed separately. The out-of-sample forecasting ability of the models is evaluated by means of several appropriate statistical indicators. RÉSUMÉ Nouvelles Méthodes de Prévisions Fondées sur les Réseaux Neuronaux Appliquées l'Emploi Régional: Une Analyse des Marchés du travail dans l'Allemagne Réunifiée Dans cet article, les auteurs ont développé une série de modèles utilisant les réseaux neuronaux (RN) pour calculer des prévisions à court terme des paramètres de l'emploi, par région allemande. Les RN sont des outils statistiques modernes fondés sur des algorithmes d'apprentissage, capables de traiter de grandes quantités de données. On s'intéresse de plus en plus aux RN car ils permettent de gérer efficacement des séries de données complexes, bien que la relation fonctionnelle entre les variables dépendantes et indépendantes n'est pas définie explicitement. Cet article compare deux méthodologies fondées sur les RN. D'abord, il utilise les RN pour prévoir l'emploi régional dans les deux régions anciennement appelées Allemagne de l'Ouest et Allemagne de l'Est. Chaque modèle réalisé calcule de simples estimations des taux de croissance d'emploi pour chaque district allemand, sur une durée de 2 ans. Puis, il calcule des prévisions complémentaires, en combinant la méthodologie RN avec une analyse shift-share (ASS). Comme l'ASS a pour but d'identifier les variations relevées sur le marché local du travail, on emploie les résultats obtenus comme variables indépendantes complémentaires dans les modèles RN. Notre échantillon de données utilisé dans nos expériences se compose de 439 districts allemands. Comme les districts composant l’échantillon présentent de grandes différences en matière de taille et d'horizon temporel, les prévisions pour l'Allemagne de l'Ouest et l'Allemagne de l'Est sont calculées séparément. La capacité des modèles à établir des prévisions hors – échantillon est évaluée avec différents indicateurs statistiques appropriés. RESUMEN Nuevos métodos de redes neurales para la previsión de empleo regional: un análisis para los mercados laborales de Alemania En este documento desarrollamos una serie de modelos de redes neurales (RN) para calcular las previsiones a corto plazo de los modelos de empleo regional en Alemania. Las RN son modernas herramientas de estadísticas basadas en algoritmos de aprendizaje capaces de procesar un gran número de datos. Las RN se están popularizando cada vez más en diferentes campos porque son capaces de manejar grupos de datos complejos cuando la relación funcional entre las variables dependientes e independientes no está explícitamente especificada. En este artículo comparamos dos metodologías de RN. Primero, utilizamos las RN para pronosticar el empleo regional en Alemania del oeste y del este. Cada modelo aplicado computa por separado los cálculos de las tasas de crecimiento de empleo para cada distrito alemán, con un intervalo de previsión de 2 años. Luego se calculan las previsiones adicionales combinando la metodología de las RN con el análisis shift-share. Dado que los análisis shift-share identifican las variaciones observadas entre los distritos laborales, sus resultados se utilizan como otras variables explicatorios en los modelos de RN. El grupo de datos utilizado en nuestros experimentos abarca un panel de 439 distritos alemanes. Las previsiones para Alemania del oeste y este se computan por separado debido a las diferencias en los horizontes de tamaño y tiempo de los datos. La capacidad de previsión a partir de las muestras en los modelos es evaluada mediante varios indicadores adecuados de estadísticas.

If you experience problems downloading a file, check if you have the proper application to view it first. In case of further problems read the IDEAS help page. Note that these files are not on the IDEAS site. Please be patient as the files may be large.

File URL:
Download Restriction: Access to full text is restricted to subscribers.

As the access to this document is restricted, you may want to look for a different version under "Related research" (further below) or search for a different version of it.

Article provided by Taylor & Francis Journals in its journal Spatial Economic Analysis.

Volume (Year): 1 (2006)
Issue (Month): 1 ()
Pages: 7-30

in new window

Handle: RePEc:taf:specan:v:1:y:2006:i:1:p:7-30
Contact details of provider: Web page:

Order Information: Web:

References listed on IDEAS
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:

as in new window
  1. Uwe Blien & Katja Wolf, 2002. "Regional development of employment in eastern Germany: an analysis with an econometric analogue to shift-share techniques," Papers in Regional Science, Springer, vol. 81(3), pages 391-414.
  2. Swanson, Norman R. & White, Halbert, 1997. "Forecasting economic time series using flexible versus fixed specification and linear versus nonlinear econometric models," International Journal of Forecasting, Elsevier, vol. 13(4), pages 439-461, December.
  3. Norman R. Swanson & Halbert White, 1997. "A Model Selection Approach To Real-Time Macroeconomic Forecasting Using Linear Models And Artificial Neural Networks," The Review of Economics and Statistics, MIT Press, vol. 79(4), pages 540-550, November.
  4. Lutz Bellmann & Uwe Blien, 2001. "Wage curve analyses of establishment data from western Germany," Industrial and Labor Relations Review, ILR Review, Cornell University, ILR School, vol. 54(4), pages 851-863, July.
  5. Matías Mayor Fernández & Ana Jesús López Menéndez, 2005. "The spatial shift-share analysis - new developments and some findings for the Spanish case," ERSA conference papers ersa05p659, European Regional Science Association.
  6. Nijkamp, Peter & Reggiani, Aura & Tsang, Wai Fai, 1999. "Comparative modelling of interregional transport flows : applications to multimodal European freight transport," Serie Research Memoranda 0002, VU University Amsterdam, Faculty of Economics, Business Administration and Econometrics.
  7. James H. Stock & Mark W. Watson, 1998. "A Comparison of Linear and Nonlinear Univariate Models for Forecasting Macroeconomic Time Series," NBER Working Papers 6607, National Bureau of Economic Research, Inc.
  8. Suahasil Nazara & Geoffrey J.D. Hewings, 2004. "Spatial Structure and Taxonomy of Decomposition in Shift-Share Analysis," Growth and Change, Wiley Blackwell, vol. 35(4), pages 476-490.
  9. Longhi, Simonetta & Nijkamp, Peter & Reggiani, Aura & Blien, Uwe, 2002. "Forecasting regional labour markets in Germany: an evaluation of the performance of neural network analysis," ERSA conference papers ersa02p117, European Regional Science Association.
Full references (including those not matched with items on IDEAS)

This item is not listed on Wikipedia, on a reading list or among the top items on IDEAS.

When requesting a correction, please mention this item's handle: RePEc:taf:specan:v:1:y:2006:i:1:p:7-30. See general information about how to correct material in RePEc.

For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Michael McNulty)

If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

If references are entirely missing, you can add them using this form.

If the full references list an item that is present in RePEc, but the system did not link to it, you can help with this form.

If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your profile, as there may be some citations waiting for confirmation.

Please note that corrections may take a couple of weeks to filter through the various RePEc services.

This information is provided to you by IDEAS at the Research Division of the Federal Reserve Bank of St. Louis using RePEc data.