Macroeconomic imbalance procedure (MIP) scoreboard indicators and their predictive strength of “multidimensional crises” Cover Image

Macroeconomic imbalance procedure (MIP) scoreboard indicators and their predictive strength of “multidimensional crises”
Macroeconomic imbalance procedure (MIP) scoreboard indicators and their predictive strength of “multidimensional crises”

Author(s): Krzysztof Biegun, Jacek Karwowski
Subject(s): Labor relations, Economic policy, Government/Political systems, Economic development
Published by: Instytut Badań Gospodarczych
Keywords: macroeconomic imbalance procedure; economic crisis; multidimensional crisis; ordered probit model;

Summary/Abstract: Research background: The evaluation of the predictive strength of MIP indicators in relation to crises is extremely important for the process of coordinating the economic policies of the EU countries. MIP is one of the pillars of the economic crisis prevention procedure. Predictive power of individual indicators has not been tested before their introduction. Purpose of the article: Evaluation of the predictive strength of fourteen MIP indicators in relation to multidimensional crises in the EU countries. Methods: We used ordered probit model to test the ability of MIP indicators to correctly predict episodes of “multidimensional crises” (as defined by the authors) in the period between 2008 and 2017 in all EU Member States. Findings & Value added: We defined “multidimensional crises”, combining several negative phenomena into one limited dependent variable. This work is also novel in its application of probit regression to test the predictive strength of MIP indicators with an ordered probit model. We identified five MIP variables which were statistically significant in predicting “multidimensional crises” for all EU countries: net international investment position, nominal unit labour cost index, house price index, private sector credit flow and general government gross debt. Other variables turned out to be less important or not effective in crises prediction.

  • Issue Year: 15/2020
  • Issue No: 1
  • Page Range: 11-28
  • Page Count: 18
  • Language: English