Does Accounting for Spatial Effects Help Forecasting the Growth of Chinese Provinces?

38 Pages Posted: 4 Nov 2009

See all articles by Eric Girardin

Eric Girardin

University Aix-Marseille 2 - GREQAM

Konstantin A. Kholodilin

German Institute for Economic Research (DIW Berlin)

Date Written: October 2009

Abstract

In this paper, we make multi-step forecasts of the annual growth rates of the real GRP for each of the 31 Chinese provinces simultaneously. Beside the usual panel data models, we use panel models that explicitly account for spatial dependence between the GRP growth rates. In addition, the possibility of spatial effects being different for different groups of provinces (Interior and Coast) is allowed. We find that both pooling and accounting for spatial effects helps substantially improve the forecast performance compared to the benchmark models estimated for each of the provinces separately. It was also shown that effect of accounting for spatial dependence is even more pronounced at longer forecasting horizons (the forecast accuracy gain as measured by the root mean squared forecast error is about 8% at 1-year horizon and exceeds 25% at 13- and 14-year horizon).

Keywords: Chinese provinces, forecasting, dynamic panel model, spatial autocorrelation, group-specific spatial dependence

JEL Classification: C21, C23, C53

Suggested Citation

Girardin, Eric and Kholodilin, Konstantin A., Does Accounting for Spatial Effects Help Forecasting the Growth of Chinese Provinces? (October 2009). DIW Berlin Discussion Paper No. 938, Available at SSRN: https://ssrn.com/abstract=1499202 or http://dx.doi.org/10.2139/ssrn.1499202

Eric Girardin (Contact Author)

University Aix-Marseille 2 - GREQAM ( email )

Centre de la Vieille Charit├ę
Marseille, 13 002
France

Konstantin A. Kholodilin

German Institute for Economic Research (DIW Berlin) ( email )

Mohrenstra├če 58
Berlin, 10117
Germany

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
21
Abstract Views
533
PlumX Metrics