Micro-modelling Public Policy: A Way out of the Crisis?

‘Nothing is certain but death and taxes’ according to an old British proverb. Everything in between is rather more complex, not least public policy. Before rolling out complicated, expensive policies on large populations, decision-makers not only need reliable data but also to run simulations.

Wherever they skip modelling, governments open themselves up to countless unforeseen factors. These include aspects that may have sparked or worsened the economic and demographic crises now facing Europe.

At a recent conference in Ireland, three of our PhD fellows presented papers of some significance to our troubled times. They focused on tax benefits for mobile workers, earnings dynamics, and retirement choice modelling — summarized in the soundcloud below.

Conference Backstory
As the European City of Science 2012, Dublin hosted the ‘European Meeting of the International Microsimulation Association‘, an international conference that brought together researchers from over 30 countries presenting 116 scientific papers.

Researchers from the German Labour Economics Institute (IZA), the National Institute for Regional and Spatial Analysis, UNICEF, and the European Dynamic Microsimulation Network, all came to discuss methodological aspects of their work. In other words, how they go about simulating and modelling complex policies.

What is Microsimulation modelling?
Microsimulation modelling is a form of computer based simulation model that simulates policy, economic, social and environmental change at a micro level (household, firm, and farm).The methodology allows one to evaluate and improve the design of public policy on a computer before rolling out often costly programmes on the general population.

To some extent the methodology can be regarded as a computer based laboratory for running policy experiments. They can thus help to improve the effectiveness and efficiency of public programmes.

A commonality across the different fields here is to use computer based techniques which aim to improve the design of public policy. To escape the economic crisis, we need to focus on methodologies that can provide better policy design. This is particularly true in Ireland, where the economic contraction over the past four years resulted from a number of policy failures.

Plenary Speakers, from left: Prof. Olivier Bargain (University of Marseilles), Prof. Richard Blundell (UCL), Dr Cathal O’Donoghue (Teagasc), Dr Andreas Peichl (IZA, Bonn), Prof Raj Chetty (Harvard University), Prof. Hilmar Schneider (IZA, Bonn)

We have much to learn from the experience of the modellers present at the conference to improve the design of our public policy. In policy making, often it is the results that are most important and influential. However in order to effectively analyse policy and other changes, it is important to develop capacity.

Technical push, focal pull
Scientific developments that focus on model design are important building blocks. To avoid re-inventing the wheel, something that unfortunately many policy modellers are frequently accused of, it is important to codify and to disseminate the knowledge that is generated through our modelling research.

The field of microsimulation has to some extent been held back by a lack of focus on this aspect in recent decades and the hope is that this conference can facilitate improved learning and development. Facilitating this there are plans for a number of journal special issues.

Participants from left: Dr Cathal O’Donoghue (Teagasc, Ireland), Dr. Jinjing Li (NATSEM, Australia), Dr. Raymond Wagener (Director of the Inspection Générale de la Sécurité Sociale (IGSS) , Luxembourg), Irina Burlacu (MGSoG, Netherlands), Dr. Philippe Liegeois (CEPS/Instead, Luxembourg)

PhD fellows and alumni from the Maastricht Graduate School of Governance presented their work at the conference. Irina Burlacu (PhD candidate) discussed the effect of tax-benefit systems on the welfare of mobile earners, using static micro-simulation on synthetic data, based on the cases of Luxembourg and Belgium.

Denisa Sologon (School of Governance alumna) proposed a new approach in simulating earnings profiles, based on a sophisticated error componence structure, which draws upon the literature used to study earnings dynamics. Jinjing Li (School of Governance alumnus) gave a presentation on retirement choice modelling and simulated retirement choices, using panel data, examining the effects of reforms on retirement. Click the soundcloud below to hear more about their presentations.

by Irina Burlacu, PhD fellow, UNU-MERIT / Maastricht Graduate School of Governance