Dual Earner Couples and Inequality in Earnings: Norway 1974-2004 by Gunn Elisabeth Birkelund Arne Mastekaasa University of Oslo.

Dual Earner Couples and Inequality in Earnings:

Norway 1974-2004by

Gunn Elisabeth BirkelundArne MastekaasaUniversity of Oslo

Rising economic inequality

• Many western countries have seen rising economic inequalities:

”The growth in market income Ginis between 1980 and 2000 ranges from a 6% to 7% increase in Denmark and Italy to a 20%-plus jump in the United Kingdom, the United States, and most surprising, Sweden.” (Esping-Andersen 2007:641)

Q: Have changes over time in married women’s labour force participation had an equalizing effect on household inequalities

in earnings?

Empirical studies has produced divergent results: Esping-Andersen (2007) found

• equalizing effect of wife’s earnings in Denmark, Sweden and the US, and

• increasing inequality effect in the UK, Germany, Italy, France and Spain.

• We will describe trends in inequalities in earnings among households in Norway for the period 1974 to 2004.

• For the period 1993-2004 we will look at both earnings and income, and include both married couples and cohabiting couples (with (a) common child(ren)).

Inequality in Norway

• A Scandinavian welfare state, with smaller socio-economic differences than most Western countries due to a compressed income distribution (Atkinsen et al. 1995)

• Partly due to oil and gas exploitation from the North sea: Norway among the top countries (highest GNP) in the world (OECD).

• Yet attituinal studies show egalitarian values still strong in the population (Aalberg 2003)

Women and work in Norway

• Married women’s labour force participation was low after WWII, and started to grow in the 1970s.

• Today, 89% of all men and 82% of all women between 25-54 are gainfully employed (Statistics Norway).

• About 40% women work part-time (less than 35 hours per week).

Inequality in household earnings

• The inequality in household’s earnings is a function of the inequality in each spouse’s earnings taken separately, and of the assiciation between the two spouses earnings; the

Coefficient of Variation:

• Where subscript f refers to household, h to husband and w to wife. CV is the coefficient of variation, the correlation of husband’s and wife’s earnings, and S is the (population) mean share of husband’s or wife’s earnings, respectively of total household earnings.

whwhhwwwhhf CVCVSSCVSCVSCV 222222

Marginal effects

Thus, the impact of wife’s earnings on household earnings inequality depends on three factors:

(1) The degree of inequality in wife’s earnings

(2) The share of wife’s earnings of household earnings

(3) The correlation of wife’s and husband’s earnings

i.e., the direction and shape of wife’s earnings impact is complicated and difficult to predict

Data

• 1974-2004: Annual Labour Force Surveys with register data on earnings

• 1993-2004: Register data on complete birth cohorts 1955-1985 (Educational careers)

• Samples: LFS: All married couples with a husband 26-65 years of age

• Samples: Educational careers: Married and cohabiting couples with a husband 35-39 years of age.

Measurements

• Pretax-pretransfer income inequalities; i.e. market inequalities (but we do not have info on capital income)

• This measurement will show larger income inequalities than a post-tax-posttransfer income measurement would.

• CVs are sensitive to extreme values, thus we have excluded households with incomes in the top 0.5 (LFS) or 0.1 (Educ career data)

Figure 1. Household CV as a function of the husband - wife

earnings correlation (ρhw) and wives' share of household

earning (Sw). Earnings inequality for husbands (CVh) set at .6

and for wives (CVw) at 1.2

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1

Husband - wife correlation (ρhw)

Sw=0.5

Sw=0.4

Sw=0.3

Sw=0.2

Sw=0.1

Figure 2. Coefficient of variation for husband's earnings and household (husband plus wife) earnings 1974-2004. 95% CI

based on bootstrapped SE's (200 replications) given by dotted lines

Spouses

Husband

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

Figure 3. CV for wives' earnings 1974-2004. Married couples with occupationally active husbands

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004

Figure 4. Husband - wife earnings correlations 1974-2004 for married couples with occupationally active husbands

-0.050

0.000

0.050

0.100

0.150

0.200

1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004

Figure 5. Wives' share of household earnings 1974-2004. Married couples with occupationally active husbands

0.000

0.050

0.100

0.150

0.200

0.250

0.300

0.350

0.400

1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004

Figure 6. Hypothetical developments in household earnings inequality under different parameter values

0.300

0.350

0.400

0.450

0.500

0.550

0.600

0.650

0.700

1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004

Constant Sw

Constant CVw

Constant ρhw

Actual CVf

Table 1. Husband – wife correlations (Pearson's r with 95% CI) for earnings, schooling, and employment 1993-2004. Households with husbands 35-39 years of age and earnings greater than zero

Employment

Earnings Schooling Full time Full or part time

Corre-lation

Lower limit

Upper limit

Corre-lation

Lower limit

Upper limit

Corre-lation

Lower limit

Upper limit

Corre-lation

Lower limit

Upper limit

1993 0.050 0.044 0.056 0.493 0.488 0.498 0.085 0.079 0.091 0.121 0.115 0.127

1994 0.047 0.041 0.053 0.486 0.482 0.491 0.086 0.080 0.092 0.125 0.118 0.131

1995 0.062 0.056 0.068 0.482 0.477 0.487 0.091 0.085 0.097 0.130 0.124 0.136

1996 0.075 0.069 0.081 0.478 0.473 0.482 0.088 0.082 0.094 0.125 0.119 0.131

1997 0.077 0.071 0.083 0.476 0.471 0.480 0.092 0.085 0.098 0.124 0.118 0.130

1998 0.077 0.071 0.083 0.473 0.468 0.477 0.089 0.083 0.095 0.119 0.113 0.125

1999 0.088 0.082 0.094 0.470 0.465 0.475 0.095 0.089 0.101 0.134 0.128 0.140

2000 0.103 0.097 0.109 0.472 0.467 0.477 0.104 0.098 0.110 0.143 0.137 0.149

2001 0.108 0.102 0.114 0.475 0.470 0.479 0.089 0.083 0.095 0.127 0.121 0.133

2002 0.106 0.100 0.112 0.478 0.474 0.483 0.086 0.080 0.092 0.128 0.122 0.134

2003 0.122 0.116 0.128 0.479 0.475 0.484 0.089 0.083 0.095 0.134 0.128 0.139

2004 0.122 0.115 0.128 0.479 0.474 0.484 - - - - - -

Note: CI's based on Fisher transformation. N=80,000 to 102,000

Figure 7. Coefficient of variation for husband's and household earnings and income. Households with occupationally activee

husbands aged 35-39, 1993-2004. 95% CI's based on bootstrapped SE's (200 replications) given by dotted lines

Household earnings

Husband's earnings

Household income

Husband's income

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Figure 8. Coefficient of variation for husband's and household earnings for married and cohabiting couples. Households with

occupationally active husbands aged 35-39, 1993-2004.

0.3

0.35

0.4

0.45

0.5

0.55

0.6

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Household, married

Husband. married

Household, cohabitors

Husband, cohabitors

Conclusions 1974-2004

• Explorative and descriptive study • Only aggregate trends:• Wife’s earnings have had an equalizing effect on

household inequalities in earnings from ca 1985-2000• CV’s for wife’s earnings declined from 1974-2004 (0

earnings)• Husband-wife’s earnings correlations increased ca 1977-

1994, then trendles fluxtuations• Wife’s share of household earnings gradually increased

in whole period 1974-2004

1994-2004

• Small – and mostly insignificant – changes in correlations (Pearson’s r) between partner’s education and work hours. Increasing association over time in partner’s earnings (yet still low)

• Stable patterns of inequality in earnings (CV’s), fluctuations in CV’s for husbands’ income (more inequality)

• Small differences between cohabitors and married couples.