Sweden and Spain: two contrasting growth experiences Matilde Mas University of Valencia and Ivie

1
Sweden and Spain two contrasting growth
experiencesMatilde MasUniversity of Valencia
and Ivie   

• Productivity Conference at Saltsjöbaden
• Stockholm
• October 5-6, 2011

2
Sweden and Spain two contrasting growth
experiences

• Objectives of the paper
• Contrast the growth experiences of the two
  countries along the period 1970-2009.
• Questions
• Why is Sweden per capita income higher?
• Which are the sources of growth in the two
  countries?
• What are the characteristics of the industrial
  specialization?
• Index OVERVIEW
• GROWTH EXPERIENCES
• INDUSTRY SPECIALIZATION
• CONCLUSIONES

3
OVERVIEW
4
Overview

• Spain and Sweden per capita income gap how has
  it evolved? A useful decomposition
• GVA/P Per capita income
• GVA/H Labour Productivity
• H/L number of hour worked per employed person
• L/LF employment rate
• LF/P activity rate (in terms of total
  population)
• GVA Gross Value Added
• P total Population
• H Total hours worked
• L total employment (persons)
• LF Labour Force

5
Overview
GVA per capita and its components
Source AMECO (2011), TCB (2011), EU KLEMS (2009)
and own calculations.
6
Overview

• Swedens per capita income has always been higher
  than the EU average with this difference
  increasing in the last years of expansion.
• On the contrary, per capita income was lower in
  Spain and this difference has maintained along
  the period.
• The gap between Sweden and Spain originated in
  Swedens higher labor productivity, as well as
  its higher rate of employment and rate of
  activity.
• Spain only overtakes Sweden in the number of
  hours worked.
• Thus, Spanish workers work more hours, are less
  productive and support a higher percentage of
  dependent population.

7
Overview
Real GVA per capita decomposition. Sweden minus
Spain (GVA per capita differences 100)
Source AMECO (2011), TCB (2011), EU KLEMS (2009)
and own calculations.

• Labour productivity has been one of the main
  determinants of per capita income differences.
• The other key factor is the activity rate.
• The Spanish unemployment rate had a negative
  effect on its relative per capita income during
  the whole period.
• Thus, Spain has a serious problem in almost all
  relevant variables that affect per capita GDP.

8
Overview
Real GVA, employment (hours worked) and labour
productivity. Total economy (Annual rates of
growth in )
Source TCB (2011), EU KLEMS (2009) and own
calculations.

• During 1970-2009, Spain showed a more dynamic
  behaviour.
• Both countries faced difficulties in creating
  jobs between 1970-1995.
• In 1995-2007, Spain experienced a productivity
  slowdown while in Sweden productivity
  accelerated.
• Thus, Spain enjoyed its highest rate of labour
  productivity growth in 1970-1995 and in
  2007-2009, precisely when it was unable to create
  employment.
• For Sweden, the golden years of productivity
  growth were 1995-2007.
• During the first two years of crisis while Sweden
  opted for labour hoarding, Spain took the
  alternative path of strong labour destruction
  mainly -though not exclusively- in the over
  dimensioned construction industry.

9
TWO CONTRASTING GROWTH EXPERIENCES
10
Two contrasting growth experiences
Gross value added. 1970-1995 and
1995-2007 (percentages)
Note TOTTotal industries ELECOMElectrical
machinery, post and communication services
MaxElecTotal manufacturing, excluding
electrical OtherGOther production
DISTDistribution FINBUFinance and business,
except real estate PERSPersonal services and
NONMARNon-market services. Source EU KLEMS
(2009) and own calculations.

• Spain outperformed Sweden in terms of GVA growth
  at the aggregated level but with important
  differences among industries.
• Sweden showed a much more dynamic behaviour in
  the ICT production sectors (ELECOM).

11
Two contrasting growth experiences
Hours worked. 1970-1995 and 1995-2007 (percentages
)
Note TOTTotal industries ELECOMElectrical
machinery, post and communication services
MaxElecTotal manufacturing, excluding
electrical OtherGOther production
DISTDistribution FINBUFinance and business,
except real estate PERSPersonal services and
NONMARNon-market services. Source EU KLEMS
(2009) and own calculations.

• The ability of the Spanish economy to create new
  jobs was astonishing after 25 years of almost
  null labour creation (its destruction since the
  beginning of the current crisis is also
  astonishing).
• While Sweden destroyed employment in ELECOM,
  Spain increased it.
• The most noticeable difference is the employment
  rate of growth in the construction industry.

12
Two contrasting growth experiences
Labour productivity. 1970-1995 and
1995-2007 (percentages)
Note TOTTotal industries ELECOMElectrical
machinery, post and communication services
MaxElecTotal manufacturing, excluding
electrical OtherGOther production
DISTDistribution FINBUFinance and business,
except real estate PERSPersonal services and
NONMARNon-market services. Source EU KLEMS
(2009) and own calculations.

• Whereas in Spain labour productivity growth
  decelerated in the expansion years, in Sweden it
  accelerated.
• In all industry aggregations labour productivity
  growth was higher in Sweden than in Spain.
• The most significant differences were found in
  ELECOM, Agriculture, MaxElec and Distribution.

13
Two contrasting growth experiences
Contributions to labour productivity growth.
1995-2007 (percentages)
Note MARKTMarket economy ELECOMElectrical
machinery, post and communication services
MaxElecTotal manufacturing, excluding
electrical OtherGOther production
DISTDistribution FINBUFinance and business,
except real estate and PERSPersonal
services. Source EU KLEMS (2009) and own
calculations.
14
Two contrasting growth experiences

• In 1995-2007 labour productivity growth was very
  high in Sweden (3.3) and very low in Spain
  (0.63).
• TFP contributions were the most divergent drivers
  of growth.
• In Spain, the highest contribution came from
  non-ICT capital deepening while in Sweden was
  TFP.
• ICT capital deepening also made a positive
  contribution in both countries but with less
  intensity in Spain than in Sweden.
• Followed closely by the contribution of labour
  composition changes, especially in the Spanish
  case.

15
INDUSTRY SPECIALIZATION
16
INDUSTRY SPECIALIZATION

• Four questions
• In which country is production more diversified?
• In which country is the sectoral composition of
  output more different from the EU-15 average?
• Are the differences in productivity growth due to
  a redistribution of factors towards industries
  with higher productivity levels or/and higher
  productivity growth (structural change effect) or
  is it a consequence of an overall pattern of
  productivity growth at industrial level
  (within-industry effect)?
• Are the differences between pairs of countries
  due to a country effect (the differences occurred
  even without any difference in the industrial
  specialization) or to a total specialization
  effect (which captures the impact of the
  different specialization)?

17
INDUSTRY SPECIALIZATION
GVA sectoral dispersion (coefficient of variation
of GVA sectoral share in total)
Source EU KLEMS (2009) and own calculations.

• The dispersion of output among the different
  industries used to be less pronounced in Spain
  than in Sweden and the EU-15 aggregate.
• However, by the end of the period the three
  converged.

18
INDUSTRY SPECIALIZATION2. In which country is
the sectoral composition of output more different
from the EU-15 average? SPAIN
Index of differences in sectoral composition (GVA
percentages)
Source EU KLEMS (2009) and own calculations.
19
INDUSTRY SPECIALIZATION

• Four questions
• In which country is production more diversified?
• In which country is the sectoral composition of
  output more different from the EU-15?
• Are the differences in productivity growth due to
  a redistribution of factors towards industries
  with higher productivity levels or/and higher
  productivity growth (structural change effect) or
  is it a consequence of an overall pattern of
  productivity growth at industrial level
  (within-industry effect)?
• Are the differences between pairs of countries
  due to a country effect (the differences occurred
  even without any difference in the industrial
  specialization) or to a total specialization
  effect (which captures the impact of the
  different specialization)?

20
INDUSTRY SPECIALIZATION
Decomposition of productivity growth. Shift-share
analysis (annual average growth rates, in )

• The main source of productivity growth is always
  the within-industry effect. That is, the one
  obtained because of the internal productivity
  improvements in each industry

Source EU KLEMS (2009) and own calculations.
21
INDUSTRY SPECIALIZATION

• Four questions
• In which country is production more diversified?
• In which country is the sectoral composition of
  output more different from the EU-15?
• Are the differences in productivity growth due to
  a redistribution of factors towards industries
  with higher productivity levels or/and higher
  productivity growth (structural change effect) or
  is it a consequence of an overall pattern of
  productivity growth at industrial level
  (within-industry effect)?
• Are the differences between pairs of countries
  due to a country effect (the differences occurred
  even without any difference in the industrial
  specialization) or to a total specialization
  effect (which captures the impact of the
  different specialization)?

22
INDUSTRY SPECIALIZATION
Decomposition of productivity growth. Shift-share
analysis (percentages)

• Swedens productivity has been always higher than
  Spain and EU-15 average, while Spains has been
  lower.
• The impact of the country effect and the
  specialization effect has been changing along the
  period.

Source EU KLEMS (2009) and own calculations.
23
FINAL REMARKS

• The overall picture is that Spain lags behind
  Sweden in almost all variables.
• Spains per capita income is lower than Sweden as
  a consequence of its lower productivity growth,
  in adition to its lower employment and activity
  rates. Spain only outperforms Sweden in the
  number of hours worked by employed person.
• Concerning productivity, the problem of the
  Spanish economy is not, or is not only, the
  result of its specialization in sectors with low
  productivity gains but it is a more general
  problem that affects all industries.
• This, together with the negative MFP
  contributions, indicates a problem in the
  fuctioning of the Spanish economy.
• Its comparison with Sweden highlights the
  distance between the two countries and the
  importance of the problems faced.
• But it also indicates that a better functioning
  of the Spanish economy is feasible.