Modelling techniques are crucial for economic forecasting and, by extension, for the planning of macroeconomic policy. Unfortunately, economists have not come up with any single model that works optimally in all situations. With numerous competing models to choose from, policy makers face a daunting task deciding which ones to utilise in addressing specific economic challenges. Their job could be made easier if they were given the ability to plug planning variables into lots of different models, providing a more complete picture of possible outcomes. Thanks to the MONFISPOL research project, macroeconomic policy makers now have this comparative option at their disposal.
Led by Professor Michel Juillard at CEPREMAP (Paris), the research consortium consisted of six partner institutions from five EU Member States (France, the UK, Spain, Germany and Italy). Among other key outputs, MONFISPOL spent three years developing and refining a database that includes around 50 empirically estimated macroeconomic models. Many of the models are commonly used by economic researchers, treasury officials and central bankers for quantitative analysis of monetary and fiscal stabilisation policies. These models, however, are generally used only in an ‘insular’ fashion and differ from one another significantly in terms of economic structure, estimation methodology and parameter estimates. Such heterogeneity makes comparison difficult.
The MONFISPOL project set out to remove the obstacles preventing easy comparison of macroeconomic models. Before the MONFISPOL project got underway, large-scale systematic comparisons of the empirical implications of multiple models (and comparisons of different policies across them) had been “infrequent and costly”1, the consortium observes. These efforts typically involved different teams of researchers, each focusing on just one single model or a subset of a model. MONFISPOL succeeded in coming up with a new comparative approach that enables researchers to perform model comparisons “easily, frequently, at low cost and on a large scale”.
For policy makers, a comparative approach to macroeconomic modelling broadens their perspective, helping them to see ‘blind spots’ they might otherwise have missed. That’s because any given policy can be seen to have different effects depending on what model is used for its evaluation. From a due diligence perspective, it is therefore advisable to assure that policy measures perform reasonably well not just in one model but in several. At least in theory, the comparative approach should enable macroeconomic researchers to offer recommendations with a stronger empirical base; and that, in turn, should allow decision makers to come up with policy rules that will be regarded as more ‘robust’.
As one might expect, the methodology employed by the MONFISPOL researchers to build their comparative database is rather technical. But basically it boils down to a process of model augmentation. MONFISPOL created a common set of comparable variables, parameters, equations and shocks that were retrofitted onto the various models. The researchers say the augmentation process was designed in such a way that new models can be added to the database easily.
Anyone wanting to access the database (which uses a software package known as DYNARE) can do so by simply visiting the website http://www.macromodelbase.com and registering. Upon registration an access code is sent via email. The code enables users to log on to the website and download the software needed to launch the tools and get started with systematic model comparison.
Acknowledging that macroeconomic modelling has been heavily criticised in the wake of the global financial crisis, the researchers point out that much of the blame has been attributed to over-reliance on “a particular class of macroeconomic models”2 that did not perform sufficiently. The MONFISPOL team have taken a different approach to the challenge, one that is open to a wide variety of paradigms and reflects the need for a more ‘pluralistic’ kind of economics. The comparative database developed by MONFISPOL should go some way toward creating a more level playing field in the competition between rival macroeconomic models and help decision makers move beyond the ‘insular’ approach to macroeconomic policy making.
1 “A New Comparative Approach to Model Building and Policy Analysis: An update”, Presentation, Volker Wieland, MONFISPOL Conference. London, 5 November 2010
2 “A New Comparative Approach to Macroeconomic Modelling and Policy Analysis”, Working Paper, Volker Wieland, Tobias Cwik, Gernot J. Müller, Sebastian Schmidt and Maik Wolters. 21 August 2009
MONFISPOL – Modelling and implementation of optimal fiscal and monetary policy algorithms in multi-country econometric models (duration: 1/10/2008 – 30/9/2011). FP7 Socio-economic Sciences and Humanities, Activity 1 “Growth, employment and competitiveness in a knowledge society”, Research area 1.3 “Strengthening policy coherence and coordination in Europe". Collaborative project (small and medium scale focused research project).
Contact: Michel Juillard, firstname.lastname@example.org