The Role of Microsimulation Models in Cross-Country Research

1
The Role of Microsimulation Models in
Cross-Country Research

• Holly Sutherland
• Institute for Social and Economic Research,
  University of Essex, UK
• IMA conference 20-22 August 2007, Vienna

2
Outline

• What do we mean by cross-national research?
• How can MSM help?
• How to do cross-national MSM lessons from
  EUROMOD
• What else can multi-country MSM do?
• Challenges and a future look

3
Cross-national research

• Comparative research
• in depth comparisons for a small number of
  countries
• traditionally qualitative assessment, even if
  data are quantitative
• quantitative analysis of many countries
• traditionally macro-level
• multi-level modelling with country effects
• replication of national studies for different
  nations
• Trans-national research
• at the level of the EU or world regions
• mobility or flows between countries

4
How can microsimulation modelling help?Examples
from tax-benefit modelling

• By providing variables that do not exist in
  available micro-data e.g.
• imputation e.g. net and gross incomes (and taxes)
  
• what if? questions (counterfactuals)
• specific aspects of taxes and benefits as
  explanatory variables e.g.
• entitlement to benefits, rather than receipt
• components a particular tax credit
• current or prospective systems rather than those
  from the recent past
• indicators of work incentives
• In a way that enhances comparability across
  countries e.g.
• equivalent outputs with different inputs
• defining and calculating variables that improve
  comparisons e.g.
• net social benefits
• child contingent payments

5
Two types of What if? question

• Effects of policies on aggregate outcomes
• Particular reforms (e.g. What is the effect on
  child poverty if child benefit is doubled?)
• Policy learning across countries policy
  swapping
• Understanding the effects of tax-benefit systems
  on different populations
• Borrowing policies that seem effective in one
  country (e.g. UK WFTC)
• Modelling individual choice budget constraints
  (incomes under a range of conditions) for
• Labour market participation and supply
• Fertility, savings and other behaviours

6
How to do it

• Try it with national models side-by-side
• hardwiring of national assumptions will need to
  be dealt with (lots of work)
• national assumptions need to be identified first
• the more countries there are the harder this is
• making a start establishes the work programme to
• Build a multi-country model
• lots of work
• but also
• more comparable results
• many uses the payoff of flexibility
• some unforeseen spin-offs the payoff of
  generality

7
EUROMOD

• EUROMOD is a general purpose multi-country
  tax-benefit model unique
• It was built because of difficulties in making
  national model calculations comparable, funded by
  a series of EC FP projects (1998-2008)
• National models exist in most of the EU15 and
  some of the 12 NMS
• Typically much more flexible than national models
  but in some cases covering the tax-benefit
  systems in less detail or more selectively
• National models for national analysis
• EUROMOD for comparative/comparable multi-country
  analysis or blue skies simulations
• Has provided national tax-benefit modelling
  capacity in/for Austria, Luxembourg, Portugal and
  Greece
• Currently EU15 and working on 12 NMS (PL, HU, SI,
  EE soon) common policy years 1998, 2001, 2003
  (most), plus various other years for selected
  countries

8
The EUROMOD approach

• Comparability through flexibility (parameterise
  everything in a structured way)
• monetary policy elements, tax rates etc units of
  assessment
• interaction of policy elements eligibility
  conditions
• income definitions
• A common framework for doing equivalent things
  across countries
• a unified design
• common structure and building blocks
• disciplined input data specification
• Complexity requiring
• tools to navigate the options
• good documentation including validation
• training and support for users
• specialist developers
• Involvement of national teams in model
  construction, development and collaborative
  research applications
• Creating a user community EUROMOD is free for
  academic use
• Learning by doing . and doing again!

9
Lessons from EUROMOD persistent challenges

• Input data reference date income reference
  period
• Accounting for benefit non take-up and tax
  evasion
• Harmonisation of input data can be a hindrance
• e.g. ECHP and SILC benefit categories
• comparable outputs often require inputs that are
  different across countries.
• A tension between quality in any one country and
  comparability across countries
• Country-specific vs. common approach in practice
  the common EUROMOD language can inhibit
  transparency of each element
• Data access permission and conditions
• The management challenge!
• Updating database and policy rules. How is it
  resourced?
• A large, long-term investment
• Multi-country, multi-institution involvement
• European Commission Framework Programme funding,
  almost entirely

10
Lessons from EUROMOD success stories

• The original design was a good one and is being
  adapted to make using it easier and maintenance
  more efficient, based on experience
• 4 NMS being included more to come
• Some unanticipated spin-offs
• Comparing policies What is the same and what is
  different what is quite similar when it looks
  different and what seems to be the same but is in
  fact quite different. etc.
• Approaching generality using the EUROMOD
  framework as a template for new national models
• My (personal) satisfaction from managing a unique
  multi-national enterprise and team
• The uses of the model
• Browse the EMWPs www.iser.essex.ac.uk/msu/emod/wo
  rkingpapers/

11
Child and non-child contingent supportas of
per capita household disposable income 2003 (2001)
Source Figari F., Paulus A. and H. Sutherland
(2007), Supporting families with children through
taxes and benefits, Policy Brief prepared for the
European Commission Observatory on the Social
Situation (SSO).
12
Future plans for EUROMOD

• A version based on EU-SILC (Statistics on Income
  and Living Conditions)
• streamlined one-stop access permission for input
  data
• facilitates regular database updating, economies
  of scale in processing, input data harmonisation
• potential to simulate short-term contribution
  conditions using the panel element
• Working on incorporation of indirect taxes and
  non-cash benefits
• Working on the modelling of benefit non take-up
  and tax evasion
• Much interest in
• behavioural response modelling linked to EUROMOD
• developing applications based on panel data
• Regularly updated policy rules
• EU27
• Other countries?
• prototype models
• for comparative purposes
• A funded programme of training and user support

13
Thanks to

• Francesco Figari
• Herwig Immervoll
• Horacio Levy
• Christine Lietz

• Daniela Mantovani
• Cathal ODonoghue
• Alari Paulus

• and everyone else who has collaborated in
  building and using EUROMOD . so far.

http//www.iser.essex.ac.uk/msu/emod/
euromod_at_essex.ac.uk