Jump to ContentJump to Main Navigation
The Changing Distribution of Earnings in OECD Countries$

A.B. Atkinson

Print publication date: 2008

Print ISBN-13: 9780199532438

Published to Oxford Scholarship Online: September 2008

DOI: 10.1093/acprof:oso/9780199532438.001.0001

Sweden

Chapter:
(p.346) Q Sweden
Source:
The Changing Distribution of Earnings in OECD Countries
Author(s):

A. B. Atkinson (Contributor Webpage)

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780199532438.003.0032

Abstract and Keywords

This chapter presents tables of data on Sweden and an account of its increased earnings distribution. It shows that Sweden's experience over the period since 1960 can be summarized as stability, followed by compression, and then reversal, with a significant rise in the top decile. As in other countries, the bottom decile has retained part of the gains from compression.

Keywords:   country data, data sources, earnings distribution, stability, compression, earnings dispersion

Wage inequality declined precipitously during the 1960s and 1970s. There was a sharp reduction in overall wage dispersion…The trend of decline in wage inequality was broken in the 1980s. Wage differentials have widened along several dimensions from the mid‐1980s to the early 1990s.

Edin and Holmlund, 1995, page 307

Q.1 Introduction

The union of Sweden and Norway was dissolved in 1905. Since then Sweden has had the same geographical boundaries. Sweden joined the European Union in 1995.

Previous Coverage

Sweden is included in all three OECD compilations:

  • 1993 (Table Q.1): all workers, and male and female separately, 1974, 1981, 1984, 1986, 1988, and 1991, supplied by Per‐Anders Edin from surveys.

  • 1996 (Table Q.2): all workers, and male and female separately, 1980 to 1993, from Income Distribution Survey.

  • LMS (Table Q.3): all workers, and male and female separately, 1980 to 1998, from Income Distribution Survey.

Official Publications

The Statistical Yearbook (Statistisk Årsbok för Sverige) has published tables with the distribution of earned income, or income from employment and business income, but these cover only certain periods. Earnings data based on the Income Distribution Survey are published in Statistiska Meddelanden (see Table Q.5). Earnings distribution data from employer surveys have been published in recent years in the Statistical Yearbook of Salaries and Wages (Lönstatistisk Årsbok) and are available from the Statistics Sweden website (see Table Q.4).

Q.2 Data Sources

Lydall (1968) gives an extensive account of distributional changes in Sweden since 1920, but this is based largely on census of population tabulations that relate to total income, not earnings. According to the criteria adopted here, these data are not (p.347) acceptable. Here I draw on data tabulated according to earnings (not including self‐employment income), drawing on the Income Distribution Survey, the Pensionable Income Register, and the earnings surveys of employers.

Income Register Data

A central element in Swedish income data is the information available in the Income Registers (Inkomst‐ och Förmögenhets‐statistiken), available annually from 1968.

The data from the registers are used in the Income Distribution Survey (Inkomstfördelningsundersökningen or HINK) that is the basis for the earnings distributions in Tables Q.2 and Q.3, and for the Luxembourg Income Study (LIS) data used by Gottschalk and Joyce (1998), reproduced in Table Q.7. The Income Distribution Survey (IDS) is an annual survey of a national sample of adults that has been conducted every year since 1975. In 2000, the interviewed sample contained some 33,000 individuals. The data combine a household survey and the income data available from the income registers. The central idea is that the survey allows, for the sampled individuals, additional information to be attached to the income tax based data derived from the complete enumeration of the income statistics. The non‐response rate to the survey in 2000 was 25.5 per cent, but the retrieval of information from the registers was almost complete. The survey non‐response does not appear to have affected the total income from employment: according to the checks reported on the LIS website, the total was only 0.5 per cent different. The earnings data in the IDS are the basis for the distributions published in Statistiska Meddelanden and reproduced in Table Q.5.

The data from the Income Registers start in 1968. For this reason, Gustavsson (2004) makes use of the earlier data from the Pensionable Income Register (PIR). He has kindly made calculations based on these data available and they are reproduced in Table Q.6, allowing coverage of the 1960s. As he notes, the data are affected by top coding, and the percentage of earnings that are top‐coded increases over time (so that the top decile is not available after 1965). The data exclude those earning less than a specified amount (about 10 per cent of the sample) and those in receipt of a pension. The sample selected was also unrepresentative in that it includes only those alive in 1990. For these reasons, Gustavsson suggests that the proportion of low earners is understated, and he shows that this is indeed the case when compared with the Income Registers (Gustavsson, 2004, Figure 1). On the other hand, the data relate to annual earnings and are not restricted to full year workers. Moreover, the PIR data do not include sickness benefits.

Employer Surveys

The earnings distribution data published in recent years in the Statistical Yearbook of Salaries and Wages (Lönstatistisk Årsbok) are based on surveys of employers in the public (census) and private sectors (sample). Surveys of employers requesting data on individual employees have been carried out for many years. For example, since 1955 surveys have been conducted of the earnings of salaried employees in the mining, (p.348)

Sweden

Figure Q.1 SWEDEN upper part of the distribution 1960–2005

Sources: Table Q.4, columns C and D; Table Q.5, columns C and D; Table Q.6, columns C and D.

manufacturing, building and construction industries (published in Löner 19xx, Del 1 Industritjänstemän, handelsanställda m fl). Other surveys covered employees in hotels and restaurants, pharmacies, commercial and savings banks, and insurance. Only in recent years have these been brought together, as in Table Q.4.

Summary

The earnings distribution data for Sweden are summarized in Figures Q.1 and Q.2. These begin with the administrative data from the Pensionable Income Register; they continue with the Income Distribution Survey data, combining register and survey data, and for the most recent years there are data from the combined employer surveys.

Q.3 What Happened?

The quotation at the beginning of this chapter summarizes concisely the history of wage dispersion since 1960, but certain aspects may be elaborated. The information about earnings in Figures Q.1 and Q.2 commences in 1960. During the decade of the 1960s, the earnings distribution appears to have been broadly stable. The downward movement in the bottom decile was less than 5 per cent and the two quintiles were little changed. According to Edin and Holmlund (1995), the solidarity wage policy pursued by the Swedish trade union movement gradually came into effect in the 1950s, following the first centralized wage rounds. The impact was described by one of the architects of the policy as follows: ‘the history of wages policy is the story of the efforts on the part of a pragmatic trade union movement to transform a sophisticated (p.349)

Sweden

Figure Q.2 SWEDEN lower part of the distribution 1960–2005

Sources: Table Q.4, columns A and B; Table Q.5, columns A and B; Table Q.6, columns A and B.

ideology of equality into a reality for the labour market’ (Meidner, 1974, page 30). To the extent that a policy of ‘equal pay for equal work’ reduced differences in pay that reflected differences in profitability, pay compression may have been expected. On the other hand, there was sizeable ‘wage drift’ (the difference between actual wage increases and the industry agreements) that may have offset the compression (see Hibbs and Locking, 1996).

‘Around the middle of the 1960s, wage solidarity took a more radically egalitarian form’ (Hibbs, 1990, page 182). Using data for blue‐collar and white‐collar workers, he finds that ‘the biggest earnings compressions came between the mid‐1960s and the mid‐1970s’ (1990, page 183). A starting date of 1968 is taken by Edin and Holmlund (1995), who describe the subsequent changes as ‘dramatic’. Combining the PIR and IDS series in Figures Q.1 and Q.2, we can see that between 1968 and 1983 the bottom decile rose by 14 per cent, and the upper quintile∕quartile fell by 8.5 per cent. Both changes clearly register on the conventions adopted here, and the change for the bottom decile is significant.21

The compression was reversed after 1983, when the wage bargaining moved much more to an industry level. Between 1983 and 1996, the bottom decile fell by 8 per cent. Again linking the PIR and IDS series, we can see that this fall still left the bottom decile at a higher percentage of the median than in 1968. The changes in the quintiles were less than 5 per cent, but the top decile rose by 16 per cent between 1983 and 2000. Gustavsson (2006) has used data from the income registers to investigate changes in (p.350) the wage structure during the 1990s. He concludes that dispersion increased in both the lower and the upper halves of the distribution. The IDS data in Figure Q.2 however show the bottom decile falling between 1990 and 2004 by 4.8 per cent. When we allow for the fact that 1990 was a local peak, we have some grounds for concluding that the increased dispersion is more evident at the top. This is supported by the employer survey data in Figures Q.1 and Q.2. From 1992 to 2005, the bottom decile fell by 5.2 per cent, or just enough to register as a fall on the criteria adopted here. The top decile, over the same period, rose by 9 per cent.

Finally, we may note that the rise in the top decile either paused (employer survey data) or was reversed (IDS data) since 2000. The latter data show the Gini coefficient for earnings as falling from 22.6 per cent in 2000 to 21.1 per cent in 2004.

Summary

The experience of Sweden over the period since 1960 can be summarized as stability, followed by compression, and then reversal, with a significant rise in the top decile. As in other countries, the bottom decile has retained part of the gains from compression.

(p.351)

Table Q.1. Sweden: OECD (1993)

Column

A

B

C

D

E

F

All

Male

Female

P10

P90

P10

P90

P10

P90

1974

74

152

76

157

75

140

1981

77

154

78

168

81

137

1984

77

148

76

150

76

134

1986

74

148

76

149

75

137

1988

76

152

76

156

79

138

1991

74

154

73

157

77

140

Coverage

All

Industry

All

Age

Adult

Sex

See headings

Occupation

All

DeWnition

Gross

Period

Hourly

Limits

None

Source

OECD (1993, Table 5.2).

Original source

Level of Living Survey (LNU) for 1974 and 1981, and Household and Nonmarket Activities Survey (HUS) for 1984, 1986, and 1988

(p.352)

Table Q.2. Sweden: OECD (1996)

Column

A

B

C

D

E

F

All

Male

Female

P10

P90

P10

P90

P10

P90

1980

77

157

76

161

80

132

1981

76

155

75

156

76

135

1982

76

153

76

157

78

134

1983

77

150

77

155

79

136

1984

75

152

76

155

78

135

1985

77

159

74

158

78

136

1986

76

157

75

160

75

135

1987

75

157

75

158

75

139

1988

75

156

75

157

77

139

1989

74

157

74

160

78

138

1990

76

152

75

156

82

140

1991

74

155

74

160

78

142

1992

75

157

74

162

77

140

1993

75

159

74

162

77

140

Coverage

All

Industry

All

Age

23 and over

Sex

See headings

Occupation

All

Definition

Gross

Intensity

FT and FY

Period

Annual

Limits

None

Source

OECD (1996, Table 3.1(

Original source

Income Distribution Survey

(p.353)

Table Q.3. Sweden: OECD (LMS)

Column

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

P

Q

R

All

Male

Female

P10

P20

P25

P75

P80

P90

P10

P20

P25

P75

P80

P90

P10

P20

P25

P75

P80

P90

1975

71.8

81.5

132.3

160.7

74.4

81.6

133.7

162.2

73.5

83.5

121.7

139.2

1978

75.9

83.1

129.5

153.0

76.3

83.3

130.1

155.5

80.6

86.4

122.5

139.6

1980

77.1

84.9

129.1

156.9

76.5

83.1

133.0

161.0

80.2

86.2

116.5

131.8

1981

75.9

84.6

129.5

155.2

75.0

82.3

130.1

156.0

76.6

85.8

117.0

134.6

1982

76.3

84.0

128.5

153.3

75.7

82.2

130.6

157.1

77.6

85.4

117.6

133.7

1983

77.0

84.3

126.1

150.5

76.7

82.6

129.1

154.9

79.1

87.8

120.4

135.6

1984

75.4

83.5

128.8

152.4

76.1

82.3

129.9

154.9

77.3

86.3

120.0

134.8

1985

77.1

84.2

130.6

158.6

74.3

81.5

131.8

158.1

78.1

87.3

117.6

135.7

1986

75.5

83.6

130.5

156.7

74.6

81.7

132.3

160.5

75.3

85.5

118.9

134.7

1987

75.0

83.6

130.6

157.3

75.2

82.5

132.4

158.4

75.1

84.9

121.4

139.1

1988

74.4

82.1

130.6

156.3

74.7

81.8

131.9

157.0

77.0

85.3

122.7

138.8

1989

74.3

81.7

131.0

157.1

73.9

81.2

131.8

159.7

77.6

86.3

123.1

138.3

1990

75.8

82.1

130.0

152.3

75.4

81.7

130.2

156.3

81.7

86.8

124.1

140.3

1991

73.4

80.6

129.1

155.1

73.3

81.5

131.1

159.9

77.7

86.1

128.0

141.9

1992

74.5

81.8

130.7

157.1

73.9

81.2

131.8

162.2

76.7

86.1

124.4

139.9

1993

74.5

81.1

131.4

159.1

73.5

80.7

133.6

162.3

76.9

85.6

123.8

139.6

1994

73.6

81.6

130.9

160.9

72.8

81.7

133.6

168.7

77.3

85.6

123.2

139.5

1995

72.1

80.4

129.9

158.9

71.6

81.2

131.4

163.0

77.1

85.1

124.7

142.4

1996

71.5

80.4

131.4

162.5

71.2

80.8

134.5

168.1

74.7

82.8

123.5

142.7

1997

72.6

81.1

129.4

160.6

72.8

81.3

132.9

168.3

76.1

84.3

124.8

141.4

1998

72.9

80.9

84.1

121.7

129.7

162.0

72.7

81.8

85.0

124.5

134.4

170.6

77.0

84.5

87.2

120.0

125.6

143.9

Coverage

All

Industry

All

Age

All

Sex

See headings

Occupation

All

Definition

Gross

Intensity

FT and FY

Period

Annual

Limits

None

Source

Downloaded from OECD website December 2005

Original source

Income Distribution Survey

Note

New basis, consistent with 1991 change in tax base, and spliced onto old series prior to 1989

(p.354)

(p.355)

Table Q.4. Sweden: Data from Statistical Yearbook of Salaries and Wages

Column

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

All

Male

Female

P10

P25

P75

P90

P90∕P10

P10

P25

P75

P90

P90∕P10

P10

P25

P75

P90

P90∕P10

1992

82.1

89.6

117.9

144.8

1.77

79.2

87.5

122.2

156.3

1.98

84.3

91.3

115.0

133.1

1.58

1993

81.0

88.3

117.5

145.3

1.80

78.2

87.8

122.5

157.8

2.01

83.9

90.8

113.9

132.3

1.58

1994

81.0

88.7

118.3

146.5

1.81

78.3

86.8

121.7

158.6

2.03

83.7

90.4

114.8

133.3

1.60

1995

80.8

89.0

117.1

145.2

1.80

78.7

88.4

121.3

158.1

2.01

82.1

90.0

113.6

131.4

1.59

1996

80.7

88.4

117.4

147.7

1.83

78.8

87.9

121.2

160.6

2.04

83.6

90.4

115.1

133.6

1.60

1997

80.3

88.3

118.5

150.0

1.87

78.6

87.9

122.5

161.9

2.06

83.0

90.2

114.4

133.3

1.61

1998

79.2

87.5

119.1

151.8

1.91

77.8

87.2

122.8

166.1

2.12

82.8

89.8

115.3

135.7

1.64

1999

78.7

87.4

119.5

154.0

1.95

76.9

86.6

123.7

166.7

2.16

81.7

89.6

116.5

137.2

1.69

2000

78.5

87.3

119.9

155.2

1.98

77.3

86.6

123.2

167.5

2.17

81.2

88.8

116.5

138.8

1.71

2001

77.9

86.8

121.1

157.9

2.03

76.7

86.1

124.3

170.3

2.22

81.4

89.3

118.1

141.2

1.74

2002

77.7

86.3

121.3

157.4

2.03

76.2

85.7

123.8

168.6

2.21

80.5

88.6

117.3

141.6

1.75

2003

77.8

87.2

121.7

157.1

2.02

76.0

85.7

124.0

165.9

2.18

80.2

88.5

117.7

141.7

1.76

2004

78.6

87.1

121.4

157.6

2.01

76.4

85.8

124.0

168.0

2.20

80.4

88.4

117.6

141.2

1.76

2005

77.8

86.6

120.8

157.9

2.03

75.7

85.7

124.3

168.7

2.23

80.4

88.7

117.6

142.2

1.77

Coverage

All

Industry

All

Age

All

Sex

See headings

Occupation

All

Definition

Gross wages and salaries

Intensity

All

Period

Monthly average

Limits

None

Source

Statistics Sweden website, downloaded 3 March 2007 and information supplied by Statistics Sweden; see also Lönestatistisk årsbok 2000, Tabell 1, 2001 Tabell 1a, 2002, Tabell 1a, and 2003 Tabell 1a

(p.356)

Table Q.5. Sweden: Income Distribution Survey

Column

A

B

C

D

E

F

G

H

I

J

K

L

All

Male

Female

P10

P25

P75

P90

P10

P25

P75

P90

P10

P25

P75

P90

1975

72.5

85.3

121.5

156.0

75.6

86.1

123.4

159.4

73.6

86.9

115.6

136.2

1978

76.7

86.4

119.7

148.5

77.7

87.5

121.0

152.8

80.7

88.8

115.3

136.5

1980

77.9

87.5

119.1

152.3

77.8

87.2

121.3

158.1

80.3

88.6

111.8

128.8

1981

76.6

87.5

120.2

150.7

76.2

86.3

120.6

153.2

76.7

89.1

112.5

131.7

1982

77.1

87.1

119.4

148.9

76.9

86.8

120.4

154.3

77.7

88.8

111.8

130.8

1983

77.8

86.9

117.2

146.1

78.0

87.2

118.2

152.1

79.2

90.6

114.4

132.7

1984

76.1

86.4

118.8

148.0

77.4

86.3

120.7

152.1

77.5

89.8

114.1

131.8

1985

77.9

87.3

121.3

154.1

75.6

85.5

121.1

155.3

78.2

89.8

112.4

132.7

1986

76.2

87.2

121.0

152.2

75.9

86.2

121.7

157.7

75.4

88.4

113.2

131.8

1987

75.8

87.0

120.7

152.7

76.5

86.3

121.9

155.6

75.1

88.4

115.0

136.2

1988

75.1

85.6

120.3

151.9

76.0

86.1

122.0

154.3

77.1

88.8

116.1

135.8

1989

75.1

85.2

120.8

152.3

75.2

85.2

121.4

156.8

77.7

89.2

116.1

135.4

1990

76.1

85.2

121.7

151.6

75.8

84.8

121.1

155.4

81.8

89.2

117.7

139.2

1991

73.5

83.7

121.8

155.1

73.3

84.5

122.7

159.8

77.7

88.7

120.9

142.0

1992

74.5

84.8

122.8

156.9

73.8

83.9

123.6

162.1

76.6

88.6

118.2

139.9

1993

74.5

84.2

123.0

159.1

73.5

83.7

124.8

162.3

77.0

88.2

118.5

139.9

1994

73.6

84.6

122.9

161.0

72.8

84.6

124.7

168.7

77.4

88.1

117.8

139.6

1995

72.1

83.7

122.4

158.9

71.6

84.8

122.7

163.0

77.1

88.2

119.3

142.5

1996

71.6

83.9

122.8

162.3

71.7

84.0

125.8

168.2

74.4

85.8

118.6

143.4

1997

72.5

84.3

121.7

160.7

72.7

84.5

123.8

168.5

75.9

87.1

119.4

141.8

1998

72.9

84.1

121.8

163.0

72.3

84.8

124.7

171.0

76.8

87.1

119.7

143.9

1999

73.5

84.2

122.6

164.5

73.0

84.6

125.0

170.5

77.5

87.4

119.4

146.5

2000

72.1

84.3

125.1

169.2

71.3

83.8

127.3

174.2

74.3

86.0

119.9

147.1

2001

72.5

83.9

124.7

166.9

71.4

83.5

126.3

172.4

75.9

86.6

120.9

151.5

2002

72.4

83.6

124.2

165.5

72.0

84.0

126.5

170.2

76.2

85.6

120.9

149.9

2003

71.4

83.6

124.3

165.3

69.9

83.4

125.8

171.6

74.3

86.1

121.9

151.6

2004

72.5

83.8

124.0

162.8

71.8

83.6

126.6

170.1

75.4

85.9

120.3

147.2

Coverage

All

Industry

All

Age

Age 20–64

Sex

See headings

Occupation

All

Definition

Gross wages and salaries, including sickness allowance

Intensity

FT FY

Period

Annual

Limits

None

Source

Statistiska Meddelanden HE21SM0601, Tabeller 27, 28, and 29

Original source

Income Distribution Survey

(p.357)

(p.358)

Table Q.6. Sweden: Pensionable Income Register (Gustavsson)

Column

A

B

C

D

P10

P20

P80

P90

1960

61.4

75.9

136.5

171.0

1961

61.6

76.0

136.2

169.8

1962

61.0

76.2

136.6

170.4

1963

60.9

76.2

137.1

170.8

1964

61.1

76.6

136.8

169.0

1965

62.6

77.2

136.5

169.4

1966

61.6

76.2

136.6

1967

60.6

75.9

136.6

1968

59.5

75.0

136.9

1969

60.2

75.5

136.3

1970

60.6

76.9

134.6

1971

59.6

76.2

132.4

1972

59.6

76.9

131.8

1973

59.8

77.7

133.0

1974

62.7

78.1

131.5

1975

63.2

78.9

130.8

1976

60.8

78.7

128.1

1977

61.0

78.1

128.3

1978

60.2

78.6

128.5

1979

59.6

78.3

129.3

1980

58.2

77.7

129.4

1981

58.2

77.6

129.7

1982

57.4

78.0

129.7

1983

58.1

77.9

130.1

1984

57.7

77.9

129.7

1985

59.0

78.1

130.8

1986

58.8

77.9

131.1

1987

57.8

77.3

131.3

1988

58.3

77.2

131.2

1989

58.3

76.9

1990

58.4

76.3

Coverage

All

Industry

All

Age

26 to 53

Sex

Male

Occupation

All

Definition

Earnings from all jobs including self‐employment

Intensity

All with earnings above basic amount

Period

Annual

Limits

None

Source

Percentiles supplied by Magnus Gustavsson

Original source

Pensionable Income Register

Note

Earnings top‐coded

(p.359)

Table Q.7. Sweden: Income Distribution Survey (Gottschalk and Joyce)

Column

A

B

P10

P90

1981

73

154

1987

72

159

1992

71

164

Coverage

Heads of household

Industry

All

Age

25 to 54

Sex

Male

Occupation

All

Definition

Gross wages and salaries

Intensity

FT Year round workers

Period

Annual

Limits

None

Source

Gottschalk and Joyce, 1998, Table 1.

Original source

Luxembourg Income Study (Income Distribution Survey)

(p.360) Bibliography

Bibliography references:

Björklund, A, 1986, ‘Assessing the decline in wage dispersion in Sweden’, in The Economics of Institutions and Markets, IUI Yearbook 1986–87, IUI, Stockholm.

—— 2000, ‘Going different ways: Labour market policy in Denmark and Sweden’, in G Esping‐Andersen and M Regini, editors, Why Deregulate Labour Markets?, Oxford University Press, Oxford.

Edin, P‐A and Holmlund, B, 1995, ‘The Swedish wage structure: The rise and fall of solidarity wage policy’, in R B Freeman and L F Katz, editors, Differences and Changes in Wage Structures, University of Chicago Press, Chicago.

—— and Zetterberg, J, 1992, ‘Interindustry wage differentials: evidence from Sweden and a comparison with the United States’, American Economic Review, vol 82: 1341–9.

Gottschalk, P and Joyce, M, 1998, ‘Cross‐national differences in the rise in earnings inequality: Market and institutional factors’, Review of Economics and Statistics, vol 80: 489–502.

Gustavsson, M, 2004, ‘Trends in the transitory variance of earnings: Evidence from Sweden 1960–1990 and a comparison with the United States’, Working Paper 2004:11, Department of Economics, Uppsala Universitet.

—— 2006, ‘The evolution of the Swedish wage structure: New evidence for 1992–2001’, Applied Economics Letters, vol 13: 279–86.

—— 2007, ‘The 1990s rise in Swedish earnings inequality: Persistent or transitory?’ Applied Economics, vol 39: 25–30.

Hibbs, D A, 1990, ‘Wage dispersion and trade union action in Sweden’, in I Persson, editor, Generating Equality in the Welfare State, Norwegian University Press, Oslo.

—— and Locking, H, 1996, ‘Wage compression, wage drift and wage inflation in Sweden’, Labour Economics, vol 3: 109–41.

—— and —— 2000, ‘Wage dispersion and productive efficiency: Evidence for Sweden’, Journal of Labor Economics, vol 18: 755–82.

Meidner, R, 1974, Co‐ordination and Solidarity: An Approach to Wages Policy, Prisma, Stockholm.

Zanchi, L, 1992, ‘The inter‐industry wage structure: Empirical evidence for Germany and a comparison with the U.S. and Sweden’, EUI Working Paper ECO 92∕76.

Notes:

(21) Although it should be noted that the PIR data show a decline in the bottom decile after 1975.