Munich Personal RePEc Archive

Ferndiagnose des RWI-Konjunkturmodells

Quaas, Georg (2019): Ferndiagnose des RWI-Konjunkturmodells.

There is a more recent version of this item available.
[thumbnail of MPRA_paper_95292.pdf] PDF

Download (258kB)


The German Institute for Economic Research Berlin (DIW), the Halle Institute for Economic Research (IWH) and the Rheinisch-Westfälisches Institut Essen (RWI) have used their own models to calculate the macroeconomic effects of the measures under the coalition agreement for the 19th parliamentary term and published them in the spring report 2018 of the Joint Economic Forecast project group. A comparison with the simulations of two other econometric models, including an older version of the RWI business cycle model (KM), suggests that these effects have been overestimated. The study focuses on the results of the RWI, which least overestimates those effects. To test the overestimation hypothesis, (1) a black box method is used to show that the implicit multipliers of the simulation results are not plausible. In a further attempt (2), the problematic effects are calculated on the basis of multipliers published in 2012 from an older version of the KM and compared with those of the new version. The hypothesis can be confirmed. Finally, (3) the latest documentation of the new versions of the KM shows that the price development indices are not calculated correctly. This would explain why the simulation results of the KM have a bias. In view of the importance of the KM for the Rheinisch-Westfälisches Institut and beyond that for the project group Gemeinschaftsdiagnose, which informs the public twice a year about the expected economic development, but also because other models may also be affected by this problem, the experts should be informed and encouraged to check their own models. The public should be warned against overly optimistic forecasts. It goes without saying that the wrong calculation method, if it is still used, must be eliminated before further simulations and forecasts are published. For a publicly funded model, it would also be necessary to document which of the technically possible methods for implementing the annual overlap method has been implemented.

Available Versions of this Item

Atom RSS 1.0 RSS 2.0

Contact us: mpra@ub.uni-muenchen.de

This repository has been built using EPrints software.

MPRA is a RePEc service hosted by Logo of the University Library LMU Munich.