Transcript Diapositiva 1

From myth to reality:
Globalisation and public spending in
OECD countries revisited
Reseaerch Note
by Marius R. Busemeyer
© Moritz Buschmann; November
2009
The research problem (1)
Research question:
How does globalisation affect public
spending in OECD countries?
© Moritz Buschmann; November
2009
The research problem (2)
Two hypothesis:
1. “Efficiency thesis”: Globalisation leads to lower taxes
because decision makers fear the exodus of capital →
lower public spending (“race to the bottom”)
2. “Compensation thesis”: Globalisation leads to problems
in societies, these problems have to be compensated by
governments → higher public spending
© Moritz Buschmann; November
2009
The research problem (3)
Literature:
• Trend: empirical evidence for
“compensation thesis”
• Domestic institutions and parties matter
© Moritz Buschmann; November
2009
The research problem (4)
Criticism on results in the literature:
•
•
•
•
•
old data (until mid. 90s)
static view (level of globalisation, not change)
“efficiency thesis” too simplistic
Narrow approaches (long run and broad effect)
No distinction between eras of globalisation
(data pooling)
© Moritz Buschmann; November
2009
Research design (1)
Dependent Variables:
•
•
•
•
total public spending (% of GDP)
total public social spending (% of GDP)
total spending on social transfers (% of GDP)
Non social public spending (% of GDP)
© Moritz Buschmann; November
2009
Research design (2)
Independent Variables:
• trade openness (∅ imports/exports % of GDP)
• KOF index (realized trade, capital flows and
domestic financial trade barriers)
• Net capital transfers (% of GDP)
© Moritz Buschmann; November
2009
Research design (3)
Control Variables:
•
•
•
•
•
•
•
•
•
Unemployment
Inflation
Well-being (national income per capita)
Age distribution (share of pop. <15 + >65)
De-industrialization index
Female labour force participation
Parties (Social Democrats in government)
EU-Membership in the 90s (Maastricht Criteria)
Country size
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
Research design (4)
The Model:
• No data pooling of years
• Simple cross-sectional regression
• Dynamic model investigating differences of the
variables within countries + Error Correction
Model (ECM) to include long term effects
(lagged variables)
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
© Moritz Buschmann; November
2009
Conclusion
What has to be discussed?
•
•
•
•
“compensation thesis” outdated?
what about “efficiency thesis”?
problem of causation
methodological lessons to learn?
© Moritz Buschmann; November
2009