Presentazione di PowerPoint

1
PHARE TWINNING CZ 2002/IB SPP/03 Joint Regional
Operational Programme and Community Support
Framework Building Implementation Capacity
Regional economic modelling for dualistic
countries the case of the Mezzogiorno
2
Guido Pellegrini, University of Bologna and
Ministry of the Economy and Finance Stefania
Rossi, University of Cagliari and Ministry of the
Economy and Finance Ornella Tarola, University of
Roma La Sapienza and Ministry of the Economy
and Finance Seminar on Macroeconomic Modelling
and CSF Policy Analysis Salone UVAL, Rome, 19th
September 2002
SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
3
The econometric model for the Mezzogiorno (MOMEZ)

• The econometric Model for the Mezzogiorno
  (MOMEZ) has been built in the Department for
  Development and Cohesion Policies, as an
  instrument for the evaluation of the impact of
  the PSM (Mezzogiorno Development Plan, now QCS)
  on growth of Southern regions in Italy

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
4
The old version of MOMEZ

• A former version of MOMEZ (2000) has been used
• for the macroeconomic ex ante evaluation of the
  PSM and the impact of CSF
• for testing the consistence between the policy
  targets and macroeconomic constraints.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
5
The new model (2002)

• The new model preserves the basic features of the
  former one
• The model captures the main growth mechanisms of
  the developing areas of a dualistic economy and
  selects the principal transmission channels of
  public policies.
• It is an instrument for policy evaluation, built
  up to give a quantitative appraisal of the
  effects of externalities generated by the QCS on
  the accumulation process in the Mezzogiorno.
• The model is focused on the supply side.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
6
New features of MOMEZ

• sectoral disaggregation
• different multipliers of heterogeneous types of
  public expenditure (infrastructures, state aids)
• an aggregate consumption function estimation
• the growth mechanism - based on the accumulation
  of physical capital - is enriched with a new
  specification, involving human capital and
  complementarities between physical capital and
  labour.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
7
This is a provisional version

• The new model is still work in progress and
  the version presented here is largely
  provisional
• the specification of some equations is not
  complete (especially for the labour market)
• other multipliers have to be analysed
• the whole set of scenarios has still to be
  built.
• So, comments are welcome! 

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
8
INDEX OF THE PRESENTATION

• Today we focus on
• Issues in setting a regional model for evaluation
  in a dualistic economy
• The theoretical framework and the specification
  of some equations
• Some very preliminary results about the main
  multipliers and elasticities.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
9
MOMEZ and QCS

• MOMEZ is strictly linked to the QCS and its
  logic.
• The basic idea of the QCS (and of the model)
  is that growth comes from the effects of actions
  on the economic and social contest, that lead to
  an increase in productivity. This induces the
  accumulation of human and physical capital, and
  increases output.
• Changes in expectations and in the social and
  economic environment can be captured by some
  specific indicators that are named in QCS and in
  MOMEZ as break-trough variables.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
10
The break trough variables (1)

• These variables have been chosen as they appear
  to have improved significantly in areas in which
  signals of remarkable local development appeared
  after the end of the old regional policies in
  Italy.
• In MOMEZ, the exogenous break-trough variables
  are included in the productivity estimation, in
  order to capture expected radical changes in
  agents behaviour and expectations.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
11
The break trough variables (2)

• They are endogenous to the model (4 exports, net
  imports, investments, participation rate) or
  exogenous (9 tourist inflow, direct investment
  from abroad, irregular employment, social
  services employment, specialized good exports,
  technological specialization, business services
  employment, interest rate gap with the North,
  crime )

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
12
The break trough variables (3)

• In MOMEZ, the growth pattern of the exogenous
  break-trough variables is estimated by
• Experts forecasts
• Based on CSF allocation
• Using the catching up with the Northern regions
  assumption

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
13
Issues in model setting

• Estimating a regional econometric model for
  evaluation in a dualistic economy requires to
  deal with several issues
• statistical issues
• theoretical issues
• econometric issues.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
14
Statistical problems (1)

• Main problem the availability of detailed
  regional data at high frequencies. The new SEC
  data cannot fulfil all the needs
• NUTS2 data are related basically to regional
  accounts at annual frequencies
• For some variables, like investment, the location
  is determined by the ownership rather than the
  use.
• Several structural territorial information are
  available only at Census data.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
15
Statistical problems (2)

• We need also sectoral disaggregated data
  sectoral patterns and transmission channels are
  different in a take off
• Therefore, we built a quite detailed sectoral
  data base, disaggregated in 4 sectors

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
16
Theoretical framework (1)

• A theoretically founded model should include an
  accumulation equation related to different
  production factors.
• A developing area suffers not only shortage of
  physical capital, but also a mismatch in human
  capital the stronger the complementarity
  between physical and human capital the wider the
  gap. This can explain why the relevant production
  constraints come from the supply side rather than
  the demand side.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
17
Theoretical framework (2)

• In the medium-long run the crucial bottleneck in
  growth is the endowment of production factors.
• The model is designed also to capture the effects
  of QCS intervention, that are mainly in areas
  such as infrastructure, human capital, and
  productivity.
• The choice is therefore in favour of a
  supply-oriented model, with a central role played
  by accumulation mechanisms.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
18
Theoretical framework (3)

• The relevance of the demand factors should not be
  underestimated, mainly in the short run.
• The central question is a peripheral area, with
  a poor endowment of physical and social capital,
  could develop its demand or rather its resources
  could be absorbed by richer, core areas ?
• Answer the South and its captive markets are
  too large and geographically not too open to be
  simply depleted from outside, even if temporary
  disequilibria have to be considered.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
19
Econometric issues (1)

• Big issue the econometric way of modelling the
  economic take off of a group of adjacent regions.
  
• If a local economy takes off, the behaviour and
  the expectations of agents in the area change
  radically with respect to the past.
• Therefore, the parameters of a model describing
  the previous behaviour of the economy will also
  change. Technically, here the Lucas critique
  clearly applies. And data on structural changes
  are not available.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
20
Econometric issues (2)

• There are two ways to tackle this problem
• to follow Lucas critique and identify the deep
  parameters modelling expectations in an explicit
  way
• To use a calibration technique which relies on
  parameters estimated in a different context.
• In our analysis we employ both of these
  approaches.

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
21
Econometric issues (3)

• We include break-trough variables in production
  equations and several break-trough variables in
  TFP equation, assuming that changes in
  expectations and behaviours of agents in the
  South - captured by those variables - can
  positively affect productivity and output

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
22
Econometric issues (4)

• For other break-trough variables, for which
  time series were not available, we have
  calibrated the coefficients in the productivity
  equation (TFP) on the basis of external
  information (mainly cross section).

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
23
WHAT THE MODEL IS NOT

•     It is not a tool designed for forecasting,
  because it is not oriented to the short term
  analysis
•  
•   It is not a tool built up to analyze the
  Southern economy, because not all the
  transmission channels are taken into account.
  Overall, the model has not a price circle,
  because our assumption is that this adjustment
  mechanism is not so important in the Mezzogiorno

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
24
MULTIPLIER ANALYSIS

• The true test of the model is in the
  evaluation of the impact of several shocks.
• We use a standard procedure, shocking some
  exogenous variables and comparing the results
  with a baseline solution. This procedure allows
  us to empirically calculate the multipliers and
  to show the long run properties of the model.
• This is basically a test of theoretical and
  empirical consistence of the model.
•  
• .

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
25
MULTIPLIER ANALYSIS 2

• . The variables are shocked starting from the
  2002. In order to compensate for a poor fit of
  the equations near the end of the time series,
  we use add factors - for the last 5 years of the
  series - in all behavioural equations, both in
  the baseline and the alternative scenario.
• Here we present the effects of the shocks on
  GDP, employment, trade balance and investment in
  the Mezzogiorno
•  
• .

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
26
Shock on world demand

• ..

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
27
Shock on North GDP

• ..

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
28
Shock on external demand

• ..

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
29
Shock on public investment

• ..

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
30
Shock on state aids

• ..

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
31
Shock on policy variables

• ..

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS
32
Shock on TFP

• ..

SEMINAR ON MACROECONOMIC MODELLING AND CSF POLICY
ANALYSIS