Dejan Krusec

Document Sample
Dejan Krusec Powered By Docstoc
					              Dejan Krušec
              PhD Student
     European University Institute
         Rethymno, 28. 8. 2003
Outline of research in the area of EMU –
    Current state and future prospects
      (At least) Three topics of interest:

1. Fiscal and monetary policy mix in the EMU
      and accession (i.e. acceding) countries
2. Exchange rate and inflation in accession
 3. “Equilibrium” level of deficit in EMU and
               accession countries
Fiscal and monetary policy mix in the EMU
     and accession (i.e. acceding) countries

Idea: What if ECB triggers with its e.g. tight monetary policy
     shocks loose fiscal policy in Eurozone countries and
     therefore countries get into the risk of exceeding the 3%
     deficit bound? Even worse scenario if some countries
     loosen and other tighten their fiscal policy as a response.
- important question for EMU: do government in response to
     monetary tightening tighten or loosen fiscal policy and
     vice versa – implications for the SGP, same question for
     accession countries,
- distinguish comovement of the two policies over the
     business cycle from reaction of one the shock of the
     other – VAR suitable technique.
   -       Research background: not much VAR method research in this
    area, some cross-section studies, e.g.
    Studies of Melitz (1997) and Wyplosz (1999) find that monetary
    and fiscal policy tend to move in the opposite directions. Melitz
    (1997, 5) uses in his study two-stage least squares and three-
    stage least squares method the interaction on the sample of 19
    OECD countries. Both authors: evidence of strategic
    substitutability between the two government policies i.e. looser
    fiscal policy promotes tighter monetary policy and vice versa.
    According to these studies we would e.g. expect the
    government spending to increase and taxation to decrease in
    response to tighter monetary policy i.e. in response to the
    increase in the interest rate - central bank’s instrument.
Why VAR approach: better isolation of shocks and responses
  (because of difference between three types of e.g. fiscal
  shocks and responses;
 -          first the automatic response of taxes and
  government spending to innovations in output, prices
  and interest rates,
 -      then the discretionary response of fiscal policy
  to output prices and interest rates and finally
 -     the structural fiscal shocks, which are unlike the
  reduced form residuals not correlated with each other
  and with all other structural shocks.
  - My previous research on fiscal-monetary policy
  mix for four OECD countries (Australia, Great
  Britain, Canada and USA) in a structural VAR
- Method: structural VAR (five variables: output,
  inflation government spending, taxation, central
  bank‟s interest rate), with the identification
  procedure following Blanchard and Perotti (QJE
3-step identification procedure

   -         first step in the identification procedure -
    fiscal variables do not react to unexpected price or
    output movements within a quarter which sets the
    discretionary response of government spending
    and taxes to movement and output to zero.
   -      second step – automatic responses: compute
    price and output elasticities of taxes, and price
    elasticity of government spending,
   -          third step: - determine which of the two
    fiscal shocks comes first – Cholesky ordering,
    determine order of other variables.
Table: Synthetic presentation of results

 Response           Australia                  Canada            Great Britain               USA
 of... to...

               Whole    To      From   Whole      To    From   Whole    To    From   Whole    To    From
                       1980     1980             1980   1980           1980   1980           1980   1980

  Gt t o        +       +        +      +         +      +      +       0        +    0       0      -
   Tt to        0       0        0       -        +      -      +       +        -     -      0      -
  Gt t o       +      na       na      +        na     na      +      na     na      0      na     na
    I t
  T t t o      0      na       na       -       na     na      +      na     na       -     na     na
    I t
  Gt t o        0       +        0      +         +      +      0       -        -    0       0      0
    rI t
   Tt to        0       0        0      0         +      -      +       +        0    0       0      -
    rI t
Countries do not have the same fiscal-monetary policy
  mix, e.g.
 -       fiscal and monetary policy acted as substitutes
  in Australia, Canada and partly in United States,
 -      the degree of substitutability is not higher in the
  period from 60s up to 1980 than after that date (from
  the 80s to the year 2001),
 -           Great Britain acts as an ‘outlier’ since the
  evidence shows that fiscal and monetary policy acted
  more like strategic complements, especially in the
  period before 1980.
Alternative approach: five variable co-
integrated VAR.
1. Suitable framework to separate automatic responses of fiscal, monetary
   policy to economic activity from fiscal and monetary policy shocks
   (i.e. separate comovement over the business cycle from pure reactions
   of the two policies to each-other). We would get three co-integrating
   relations, one for automatic response of government spending to GDP
   and inflation movements, the other for automatic response of taxation
   to GDP and inflation movements and the money demand equation.
2. No need to get disaggregated data on taxes and government spending
   to compute price and output elasticities
3. It is possible to determine which of the two policies comes first i.e. is
   weakly exogenous with respect to the other and with respect to output
   and price movements.
4. Cholesky ordering of variables needed only to identify the shocks once
   cointegration is accounted for.
Same method for accession countries, comparison of
  responses, implication for ECB.
Same method applied ot Eurozone as a whole
  with the aggregation like in e.g. Artis and
  Beyer (2003?), but the question of
  usefullness of such aggregation since no
  SGP before 1999, therefore variables
  combine differently then after 1999.
2. Exchange rate and inflation in accession

- Well known question of how to enter into the ERM II and
    whether or not the Maastricht inflation criterium is too
-   Two potential problems for accession countries:
1.  Balassa-Samuelson effect (driven by productivity)
2.  Exchange rate pass-through (driven by inflation
    expectations in a small open economy)

Existing studies analyse both effects separately, but no
     analysis of both in the same framework.
Proposed method
Analyse both effects together in a co-integrated VAR framework, where
   we have four variables:
   - productivity in the tradable sector,
  - money supply,
  - exchange rate,
  - inflation rate (or difference of domestic and Eurozone inflation rate).
We expect to have two cointegrating relationships, since we have a policy
   variable and an exogenous variable and two endogenous variables.
Existing study of Coricelli et al. (2002) on exhange rate pass-through does
   not take into account the Balassa-Samuelson effect.
First cointegrating relation (automatic response
productivity combined with exchange rate and
  inflation rate – Balassa Samuelson effect, LR test
  of significance of parameteres,
Second cointegrating relation:
Exchange rate pass through - money supply
  combined with exchange rate and inflation rate
  (and LR test).
Possibilities of the method
Checking for interdependence of the two effects in
   the direction that the productivity growth shock
   may trigger the money supply growth and
   exchange rate adjustment which in turn induces
If this is empirically true, productivity effects
   inflation directly and through the monetary policy
3. Equilibrium level of deficit in EMU and
accession countries
Idea: What if EMU and accession countries have
  different „equilibrium“ levels of deficit, what if
  for some countries it is nested in institutions that
  they have higher budget deficits in absence of any
  shocks, while others run balanced budgets ór even
  budget surpluses?
- Scenario: High „equilibrium“ deficit + recession
  or negative supply or demand shock could mean
  exceeding the 3% deficit bound
   Could work in the same cointegrated VAR
    framework with variables included:
-   GDP,
-   Inflation,
-   CB„s interest rate,
-   deficit.
The framework allows to do the LR test whether deficit
  „equilibrium“ response to the GDP and inflation
  movements are significantly different between the
  countries. Additionally, we can test for responses of the
  deficit to the shocks in the CB„s interest rate.

Shared By: