Extract from Divided We Stand: Why Inequality Keeps Rising

Transcription

1 Extract from Divided We Stand: Why Inequality Keeps Rising (2011) James J. Heckman University of Chicago AEA Continuing Education Program ASSA Course: Microeconomics of Life Course Inequality San Francisco, CA, January 5-7, 2016

2 AN OVERVIEW OF GROWING INCOME INEQUALITIES IN COUNTRIES: MAIN FINDING Table 1. Household incomes increased faster at the top Table 1. Household incomes increased faster at the top Trends in real household income by income group, mid-1980s to late 2000s Average annual change, in percentages Total population Bottom decile Top decile Australia Austria Belgium Canada Chile Czech Republic Denmark Finland France Germany Greece Hungary Ireland Israel Italy Note: Japan Income refers to disposable household income, 0.3 corrected for household 0.5 size and deflated by the consumer 0.3 price index (CPI). Luxembourg Average annual changes are calculated over 2.2 the period from 1985 to 2008, 1.5 with a number of exceptions: was the earliest year for Austria, Belgium, and Sweden; 1984 for France, Italy, Mexico, and the United States; 1986 for Finland, Mexico Luxembourg, and Norway; 1987 for Ireland; 1988 for Greece; 1991 for Hungary; 1992 for the Czech Republic; and 1995 for Australia Netherlands and Portugal. The latest year for Chile 1.4 was 2009; for Denmark, Hungary, 0.5 and Turkey it was 2007; 1.6 and for Japan New Changes Zealand exclude the years 2000 to for Austria, Belgium, Ireland, 1.1 Portugal and Spain for which 2.5 surveys were not comparable. Norway Information Portugal on data for Israel: Source: Database on Household Income Distribution and Poverty. Spain

3 Denmark Finland AN OVERVIEW 1.7 OF GROWING INCOME 1.2 INEQUALITIES IN COUNTRIES: 2.5 MAIN FINDING France Table 1. Household incomes increased faster at the top Table 1. Household incomes increased faster at the top Trends in real household income by income group, mid-1980s to late 2000s Germany Greece Hungary Ireland 3.6 Average annual change, 3.9 in percentages 2.5 Israel Italy Total population 0.8 Bottom 0.2 decile Top 1.1 decile Japan Australia Luxembourg Austria Mexico Belgium Netherlands Canada New Zealand Chile Norway Czech Republic Portugal Denmark Spain Finland Sweden France Turkey Germany United Kingdom Greece United States Hungary Ireland Israel Note: Italy Income refers to disposable household 0.8 income, corrected for household 0.2 size and deflated by 1.1 the consumer Note: price Income index refers (CPI). toaverage disposable annual household changes income, are calculated corrected for over household the period size and fromdeflated 1985 toby 2008, the consumer with a number price index of Japan (CPI). exceptions: Average 1983 annual was changes the earliest are calculated year for over Austria, the period Belgium, from and 1985Sweden; to 2008, 1984 with for a number France, ofitaly, exceptions: Mexico, 1983 and was the the earliest Luxembourg United year States; for Austria, 1986 for Belgium, Finland, andluxembourg, Sweden; 1984and for France, Norway; Italy, 1987 Mexico, for Ireland; and the1988 United for States; Greece; for for Finland, Hungary; Luxembourg, Mexico for the and Czech Norway; Republic; 1987 for and Ireland; 1995 for 1988 Australia for Greece; and 1991 Portugal. for Hungary; The latest 1992 year for for the Chile Czech was Republic; 2009; for and Denmark, 1995 for Australia Hungary, Netherlands Portugal. The latest year for Chile and Turkey it was 2007; and for 1.4was 2009; for Denmark, Hungary, Japan Changes exclude 0.5 and Turkey it was 2007; the years 2000 to 2004 for Austria, 1.6and for Japan Belgium, Ireland, New Changes Zealand exclude the years 2000 to 2004 Portugal and Spain for which surveys 1.5 for Austria, Belgium, Ireland, were not comparable. 1.1 Portugal and Spain for which 2.5 surveys were not comparable. 1. Norway Information on data for Israel: Information on data for Israel: Source: Source: Portugal Database on Household Income2.0 Distributionand andpoverty Spain

4 Figure 1. Income inequality increased in most, but not all countries AN OVERVIEW OF GROWING INCOME INEQUALITIES IN COUNTRIES: MAIN FINDINGS Figure 1. Income inequality increased in most, but not all countries Gini coefficients of income inequality, mid-1980s and late 2000s ( ) 0.50 Increasing inequality Little change in inequality Decreasing inequality Mexico United States Israel 1 United Kingdom Italy Australia New Zealand Japan Canada Germany Netherlands Luxembourg Finland Sweden Czech Republic Norway Denmark Note: For data years see Table 1. Little change in inequality refers to changes of less than 2 percentage points. 1. Information on data for Israel: Note: Source: Income refers Database toon disposable Household Income household Distribution income. and Poverty. For data years see Table 1. Little change in inequality refers to changes of less than 2 percentage points Information on data for Israel: Source: Database on Household Income Distribution and Poverty. the major underlying, indirect causes of changes in inequality. Is globalisation the main culprit? To what degree were changes in labour and product Divided market We Stand policies Extract and France Hungary Belgium Turkey Greece

5 On the other hand, most policy and institutional reforms also contributed to widening wage disparities, as more low-paid people entered employment and the highly skilled Figure 4. Product and labour market regulations and institutions became weaker Figure 4. Product and labour market regulations and institutions became weaker Developments in product market regulation, employment protection legislation, tax wedges and union density, average, (1980 = 100) 150 PMR EPL Tax wedge Union density = = Note: PMR is a summary indicator for product market regulation. EPL is a summary indicator of the strictness of overall employment protection legislation (only available from 1985 onwards). Tax wedge refers to an average Note: worker PMR and is is athe summary of indicator income tax for product and employees market regulation. and employers EPL payroll is a summary taxes as a indicator percentage of the of labour strictness costs. of overall employment Union density protection is the legislation number (only of union available members fromas 1985 a proportion onwards). of Tax all employees wedge refers eligible to to anbe average members. worker and is the sum of income tax and employees and employers payroll taxes as a percentage of labour costs. Union density is the number Source: See Chapter 1, Figure of union members as a proportion of all employees eligible to be members. Source: See Chapter 1, Figure 1.18.

6 Figure 5. Levels of earnings inequality are much higher when part-timers and self-employed are accounted for Earnings inequality (Gini coefficients) among full-timers, part-timers and all workers including the self-employed, mid-2000s 0.50 Full-time wage workers Full-time and part-time wage workers All workers including self-employment ( ) Countries reporting gross earnings Countries reporting net earnings Denmark Czech Republic n.a. Sweden Finland n.a. Australia Norway Netherlands Germany United Kingdom Israel 1 Canada United States Average 0.20 Belgium Hungary Italy Spain Austria Greece France Ireland Luxembourg Poland Mexico Average Note: Data presented on the individual level. Working-age individuals living in a working household. Countries are presented in increasing order of earnings inequality among all workers. Data refer to a year between 2003 and 2005, except for Belgium and France (2000). 1. Information on data for Israel: Source: Chapter 4, Figure 4.1.

7 Figure 6. Hours worked declined more among lower-wage workers N OVERVIEW OF GROWING INCOME INEQUALITIES IN COUNTRIES: MAIN FINDINGS Figure 6. Hours worked declined more among lower-wage workers Trends in annual hours worked by the bottom and top 20% of earners, average, mid-1980s to mid-2000s Top quintile Bottom quintile Total Percentage change in hours worked Note: Paid workers of working age. Source: Chapter 4, Figure 4.5. Note: Data presented on the individual level. Paid workers of working 1 age. 2 Source: Chapter 4, Figure 4.5. These trends contributed to higher household earnings inequality in the period under study. Some observers even consider changes in family formation to be the main reason for rising inequality. Daly and Valletta (2006), for instance, suggest Divided that We the Stand increase Extractin single-

8 Figure 7. Demographic changes were less important than labour market trends in explaining changes in household earnings distribution Percentage contributions to changes in household earnings inequality, average, mid-1980s to mid-2000s Men s earnings disparity Assortative mating Residual -19% 42% 17% 11% 11% 39% Women s employment Men s employment Household structure Percentage contribution Note: Working-age population living in a household with a working-age head. Household earnings are calculated as the sum of earnings from all household members, corrected for differences in household size with an equivalence scale (square root of household size). Percentage contributions of estimated factors were calculated with a decomposition method which relies on the imposition of specific counterfactuals such as: What would the distribution of earnings have been in recent year if workers attributes had remained at their early year level? The residual indicates the importance of unmeasured factors. These include other changes in household characteristics, such as trends in ageing or migration. Source: Chapter 5, Figure 5.9.

9 Figure 9. Market incomes are distributed much more unequally than net incomes Inequality (Gini coefficient) of market income and disposable (net) income in the area, working-age persons, late 2000s Slovenia Denmark Czech Republic Gini coefficient of market income Gini coefficient of disposable income ( ) Slovak Republic Norway Belgium Finland Sweden Austria Hungary Ireland Switzerland Luxembourg France Netherlands Germany Korea Iceland Estonia Greece Poland Spain New Zealand Japan Australia Canada Italy United Kingdom Portugal Israel 1 United States Turkey Mexico Chile Note: Income refers to household income. Late 2000s refers to a year between 2006 and The average excludes Greece, Hungary, Ireland, Mexico and Turkey (no information on market income available). Working age is defined as years old. Countries are ranked in increasing order of disposable income inequality. 1. Information on data for Israel: Source: Chapter 6, Figure

10 Figure 10. While market income inequality rose, redistribution through tax/transfers became less effective in many countries Changes in cash redistribution of social transfers, personal income taxes and social security contributions, mid-1980s to mid-2000s Beginning of period Middle of period End of period 14 A. Social transfers 14 B. Personal income tax AUS 85/95/03 CAN 87/94/04 CZE 92/96/04 DNK 87/95/04 FIN 87/95/04 DEU DEU-W 84/94/04 ISR 1 86/97/05 NLD 83/94/99 NOR 86/95/04 POL SWE 87/95/05 CHE 82/92/04 GBR 86/94/95/04 USA 86/94/ AUS 85/95/03 CAN 87/94/04 CZE 92/96/04 DNK 87/95/04 FIN 87/95/04 DEU DEU-W 84/94/04 ISR 1 86/97/05 NLD 83/94/99 NOR 86/95/04 POL SWE 87/95/05 CHE 82/92/04 GBR 86/94/95/04 USA 86/94/04 Note: Income refers to individual income. Redistribution is the difference between the Gini coefficients before and after the respective tax or benefit. Households headed by a working-age individual. 1. Information on data for Israel: Source: Chapter 7, Figure 7.3.

11 1 2 Figure 12. The share of top incomes increased, especially in English-speaking countries Figure 12. The share of top incomes increased, especially in English-speaking countries Shares of top 1% incomes in total pre-tax incomes, (or closest year) % of total pre-tax income United States United Kingdom Canada Germany Switzerland Ireland Portugal Italy Japan New Zealand Australia France Spain Finland Belgium Denmark Norway Sweden Netherlands Note: 2007 values refer to 2006 for Belgium, France and Switzerland; 2005 for Japan, Netherlands, New Zealand, Note: 2007 values refer to 2006 for Belgium, France and Switzerland; 2005 for Japan, Netherlands, New Zealand, Portugal, Portugal, Spain and the United Kingdom; 2004 for Finland; and 2000 for Germany and Ireland. Countries are ranked Spain and the United Kingdom; 2004 for Finland; and 2000 for Germany and Ireland. Countries are ranked by decreasing by decreasing shares in the latest year. shares in the latest year. Source: Source: Chapter Chapter 9, Figure 9, Figure 9.A A

12 indicates that the Gini coefficient of earnings among the whole working-age population on average decreased by 0.04 percentage points annually over the mid-1980s to mid-2000s. Figure This 3.2. is the Estimated net outcome contributions of the two of opposing wage dispersion forces: and increasing employment wage effects dispersion to overall among the earnings employed inequality has exerted amonga the disequalising working-ageimpact, population contributing 0.11 percentage point a year to Figure 3.2. Estimated contributions of wage dispersion and employment effects to overall earnings inequality among the working-age population Average annual percentage-point change in overall Gini Contribution of wage effect Contribution of employment effect Residuals Note: The contribution of each variable is computed as the average annual change in the variable multiplied by the regression coefficient (Table 3.1) on that variable. Note: Data presented on the individual level. The contribution of each variable is computed as the average annual change in the Source: variable multiplied Secretariat by the calculations regression coefficient from the (Table Luxembourg 3.1) on Income that variable. Study (LIS). Source: Secretariat calculations from the Luxembourg Income 1 Study (LIS) DIVIDED WE Divided STAND: We WHY Stand INEQUALITY Extract KEEPS RISING

13 Table 3.A2.1. Simulation of the wage and employment effects by country, entire working-age population Actual Gini coefficient of earnings Decomposition of change in Gini coefficient First year Last year Change (2)-(1) Wage inequality effect Employment effect Residuals (1) (2) (3) (4) (5) (6) Australia (85-03) Austria (94-04) Belgium (85-00) Canada (87-04) Czech Republic (92-04) Denmark (87-04) Finland (87-04) France (81-00) Germany (84-04) Greece (95-04) Hungary (91-05) Ireland (94-04) Israel (79-05) Italy (87-04) Luxemboug (85-04) Mexico (84-04) Netherlands (83-04) Norway (79-04) Poland (92-04) Spain (95-04) Sweden (81-05) Switzerland (00-04) United Kingdom (86-04) United States (79-04) Information on data for Israel: Source: Secretariat calculations from the Luxembourg Income Study (LIS).

14 Figure 4.1. Earnings inequality (Gini coefficient) among full-time workers, full-time and part-time workers and all workers, mid-2000s Full-time workers Full-time and part-time workers All workers including self-employment ( ) 0.50 Countries reporting gross earnings 0.50 Countries reporting net earnings DNK (2004) CZE (2004) n.a. SWE (2005) FIN (2004) AUS (2003) n.a. NOR (2004) NLD (2004) DEU (2004) GBR (2004) ISR (2005) 1 CAN (2004) USA (2004) Average BEL (2000) HUN (2005) ITA (2004) ESP (2004) AUT (2004) GRC (2004) FRA (2000) IRL (2004) LUX (2004) POL (2004) MEX (2004) Note: Data presented on the individual level. Samples are restricted to the civilian working-age population (25-64 years). n.a.: Not available. 1. Information on data for Israel: Source: Secretariat calculations from the Luxembourg Income Study (LIS). Average

15 any country. Figure 4.4. Figure Inequality 4.4. Inequality of hourly wages of hourly versus inequality wages versus of annual inequality earnings, all of paid annual workers earnings, all paid workers Gini coefficient of annual earnings USA ISR 2 CAN DEU GBR 1 NLD LUX FRA FIN ESP IRL AUS 1 GRC HUN AUT CZE 1 BEL ITA MEX Gini coefficient of hourly wages Note: Samples are restricted to all paid workers (aged 25-64) with positive wages/positive Note: Data presented on the individual level. Samples are restricted to all paid workers (aged 25-64) with positive wages/positive hours hours worked worked during during the the reference reference year. year. Data Data refer refer to the to year the 2004, year except 2004, for except Australia for Australia (2003), Belgium and France (2000). (2003), For Belgium Finland, and hourly France wage is (2000). calculated For Finland, based on imputed hourly wage hours is worked calculated per week. based on imputed 1. Hourlyhours wage isworked calculated per based week. on imputed weeks worked. 2. Hourly1. wage Hourly is calculated wage is based calculated on working based 52 weeks. on imputed Information weeks onworked. data for Israel: 2. Hourly wage is calculated based on working 52 weeks. Information on data for Israel: Source: Secretariat calculations from the Luxembourg Income Study (LIS).

16 II.4. HOURS WORKED, SELF-EMPLOYMENT AND JOBLESSNESS AS INGREDIENTS OF EARNINGS INEQUALIT Table 4.1. Decomposition of the variance of log annual earnings, paid workers, mid-2000s Table 4.1. Decomposition of the variance of log annual earnings, paid workers, mid-2000s Var(ln_annual earnings) Var(ln_hourly wages) Var(ln_annual hours) 2xCov(ln_hwage, ln_ahours) (1) (2) (3) (4) Panel A. Countries reporting gross earnings Australia (1.00) (0.457) (0.554) (0.011) Canada (1.00) (0.607) (0.144) (0.249) Czech Republic (1.00) (0.721) (0.132) (0.147) Finland (1.00) (0.510) (0.215) (0.275) Germany (1.00) (0.405) (0.306) (0.289) Israel (1.00) (0.655) (0.257) (0.086) Netherlands (1.00) (0.449) (0.326) (0.225) United Kingdom (1.00) (0.496) (0.327) (0.176) United States (1.00) (0.617) (0.224) (0.158) Average (0.546) (0.276) (0.177) Corr(AE, hw) = 0.91 Corr(AE, ah) = 0.43 Panel B. Countries reporting net earnings Note: Austria Data 2004 presented on the individual level. (1.00) Samples are restricted (0.726) to all paid workers (aged (0.502) 25-64) with positive (0.227) wages and positive hours worked during the reference year. For Finland, hourly wage is calculated based on imputed hours worked per Belgium (1.00) (0.584) (0.388) (0.028) week. Numbers in parentheses refer to the fraction of variance of log annual earnings. 1. Hourly France 2000 wage is calculated based on imputed (1.00) weeks worked (0.417) (0.471) (0.112) 2. Hourly Greece 2004 wage is calculated based on working (1.00) 52 weeks Information (0.723) on data for Israel: (0.434) (0.157) Hungary (1.00) (0.600) (0.313) (0.086) Source: Secretariat calculations from the Luxembourg Income Study (LIS). Ireland (1.00) (0.437) (0.563) (0.000) Italy (1.00) (0.730) Divided (0.420) We Stand Extract (0.150)

17 Czech Republic (1.00) (0.721) (0.132) (0.147) II.4. HOURS WORKED, SELF-EMPLOYMENT AND JOBLESSNESS AS INGREDIENTS OF EARNINGS INEQUALIT Finland (1.00) (0.510) (0.215) (0.275) Germany (1.00) (0.405) (0.306) (0.289) TableIsrael 4.1. Decomposition of the variance of log annual earnings, paid workers, mid-2000s Table Decomposition of (1.00) the variance of (0.655) log annual earnings, (0.257) paid workers, (0.086) Netherlands (1.00) mid-2000s (0.449) (0.326) (0.225) United Kingdom (1.00) (0.496) (0.327) (0.176) United States xCov(ln_hwage, Var(ln_annual earnings) (1.00) Var(ln_hourly wages) (0.617) Var(ln_annual (0.224) hours) (0.158) ln_ahours) Average (0.546) (0.276) (0.177) (1) Corr(AE, (2) hw) = 0.91 Corr(AE, (3) ah) = 0.43 (4) Panel A. B. Countries reporting gross net earnings Australia Austria (1.00) (0.457) (0.726) (0.554) (0.502) (0.011) (0.227) Canada Belgium (1.00) (0.607) (0.584) (0.144) (0.388) (0.249) (0.028) Czech FranceRepublic (1.00) (0.721) (0.417) (0.132) (0.471) (0.147) (0.112) Finland Greece (1.00) (0.510) (0.723) (0.215) (0.434) (0.157) (0.275) Germany Hungary (1.00) (0.405) (0.600) (0.306) (0.313) (0.289) (0.086) Israel Ireland (1.00) (0.655) (0.437) (0.257) (0.563) (0.086) (0.000) Netherlands Italy (1.00) (0.449) (0.730) (0.326) (0.420) (0.150) (0.225) United Luxembourg Kingdom (1.00) (0.496) (0.567) (0.327) (0.344) (0.176) (0.089) United MexicoStates (1.00) (0.617) (0.961) (0.224) (0.168) (0.128) (0.158) Average Spain (1.00) (0.546) (0.529) (0.276) (0.393) (0.177) (0.078) Average Corr(AE, hw) = (0.627) 0.91 Corr(AE, ah) = (0.400) (0.027) Corr(AE, hw) = 0.78 Corr(AE, ah) = 0.31 Panel B. Countries reporting net earnings Note: Samples are restricted to all paid workers (aged 25-64) with positive wages and positive hours worked during Note: the Austria Data reference 2004 presented year. on For the Finland, individual hourly level. (1.00) wage Samples is calculated are restricted based (0.726) toon all imputed paid workers hours (aged (0.502) worked 25-64) per with week. positive Numbers (0.227) wagesinand positive parentheses Belgium hours 2000worked refer to during the fraction the0.358 reference of variance year. (1.00) For of log Finland, annual hourly earnings. (0.584) wage is calculated based (0.388) on imputed hours worked (0.028) per week. Numbers in parentheses refer to the fraction of variance of log annual earnings. 1. France Hourly 2000wage is calculated based on imputed (1.00) weeks worked. (0.417) (0.471) (0.112) 1. Hourly wage is calculated based on imputed weeks worked. 2. Greece Hourly 2004wage is calculated based on (1.00) working weeks. Information (0.723) on data for Israel: (0.434) 2. Hourly is calculated based on working 52 weeks. Information on data for Israel: (0.157) Hungary (1.00) (0.600) (0.313) (0.086) Source: Source: calculations the Luxembourg Income Study (LIS). Ireland 2004 Secretariat calculations from(1.00) the Luxembourg Income (0.437) Study (LIS) (0.563) (0.000) Italy (1.00) (0.730) Divided(0.420) We Stand Extract (0.150)

18 which compares the percentage change in annual hours (left panel) and the change in hourly real wages (right panel) among workers in the bottom and top quintiles. At first Figure 4.5. glance, Changes Figurein4.5annual suggests hours that a worked decline in and low-paid in hourly workers real hours wagesis by an important earnings quintile, factor in the rise of inequality in most countries. mid-1980s to mid-2000s Figure 4.5. Changes in annual hours worked and in hourly real wages by earnings quintile, mid-1980s to mid-2000s Changes in annual hours NLD DEU Bottom quintile Top quintile Panel A. Countries reporting gross earnings Changes in hourly wages ISR CZE GBR CAN USA FIN AUS Average % Panel B. Countries reporting net earnings Changes in annual hours Changes in hourly wages Note: Data presented on the individual level. Samples are restricted to all paid workers (aged 25-64) with positive wages and LUX positive hours worked during the reference year with information on annual hours worked. Mean wages in national currencies at constant ITA values. Countries ranked in descending order of changes in earnings inequality (see Table 4.A1.2). 1. Information MEX on data for Israel: Source: BEL Secretariat calculations from the Luxembourg Income Study (LIS). GRC 95-04

19 CAN USA Figure 4.5. Changes in annual hours worked and in hourly real wages by earnings quintile, FIN mid-1980s AUS 85-03to mid-2000s Average % Changes in annual hours LUX ITA MEX BEL GRC AUT FRA IRL ESP HUN Average Panel B. Countries reporting net earnings Changes in hourly wages % Note: Samples are restricted to all paid workers (aged 25-64) with positive wages and positive hours worked during the reference year with information on annual hours worked. Mean wages in national currencies at constant 2005 values. Countries ranked in descending Note: order Data of changes presented in earnings on the inequality individual (seelevel. TableSamples 4.A1.2). are restricted to all paid workers (aged 25-64) with positive wages and positive 1. Information hours worked on data during for Israel: thehttp://dx.doi.org/ / reference year with information on annual hours worked. Mean wages in national currencies atsource: constant 2005 Secretariat values. calculations Countries from ranked the inluxembourg descendingincome order Study of changes (LIS). in earnings inequality (see Table 4.A1.2). 1. Information on data for Israel: Source: Secretariat calculations from the Luxembourg Income Study (LIS). Among the countries reporting gross earnings, both changes in hours and hourly wage

20 Figure 5.1. Inequality (Gini coefficient) of annual earnings among individuals and households, all working-age II.5. TRENDS households IN HOUSEHOLD (including EARNINGS INEQUALITY: individuals THE ROLE and OF CHANGING households FAMILY FORMATION with no PRACTICES earnings) Figure 5.1. Inequality (Gini coefficient) of annual earnings among individuals and households, all working-age households (including individuals and households with no earnings) Individual earnings ( ) Household earnings per earner Equivalent household earnings 0.7 Countries reporting gross earnings 0.7 Countries reporting net earnings DNK (2004) SWE (2005) NOR (2004) FIN (2004) CZE (2004) NLD (2004) DEU (2004) AUS (2003) CAN (2004) GBR (2004) USA (2004) ISR (2005) 1 Average Note: Samples are restricted to the working-age population (25-64 years) living in a household with a working-age head. Estimates include individuals and households with no earnings. Equivalent household earnings are calculated as the sum of earnings from all Note: Samples household are members, restricted corrected to the for working-age differences household population size (25-64 with an equivalence years) living scale in(square a household root of household with a working-age size). head. 1. Information on data for Israel: Estimates include individuals and households with no earnings. Equivalent household earnings are calculated as the sum of Source: Secretariat calculations from the Luxembourg Income Study (LIS). earnings from all household members, corrected for differences in household size with an equivalence scale (square root of household size). 1. Information on data for Israel: Source: Secretariat calculations from the Luxembourg Income Study (LIS). Figure 5.2. Inequality (Gini coefficient) of annual earnings among individuals and households, workers and working households 0.2 FRA (2000) AUT (2004) ESP (2004) LUX (2004) BEL (2000) ITA (2004) IRL (2004) HUN (2005) GRC (2004) POL (2004) MEX (2004) Average

21 Figure 5.2. Inequality (Gini coefficient) of annual earnings among individuals and households, workers and working households Individual earnings ( ) Household earnings per earner Equivalent household earnings 0.50 Countries reporting gross earnings 0.50 Countries reporting net earnings DNK (2004) CZE (2004) SWE (2005) FIN (2004) AUS (2003) NOR (2004) NLD (2004) DEU (2004) GBR (2004) ISR (2005) 1 CAN (2004) USA (2004) Average BEL (2000) HUN (2005) ITA (2004) ESP (2004) GRC (2004) AUT (2004) FRA (2000) LUX (2004) IRL (2004) POL (2004) MEX (2004) Note: Samples are restricted to the working-age population (25-64 years) living in a household with a working-age head and positive earnings. Equivalent household earnings are calculated as the sum of earnings from all household members, corrected for differences in household size with an equivalence scale (square root of household size). 1. Information on data for Israel: Source: Secretariat calculations from the Luxembourg Income Study (LIS). Average

22 2 0 Figure 5.7. Degree of assortative mating, stricter and broader definitions DNK (87-04) SWE (81-05) Panel B. Percentage of workers in earnings quintile i with a spouse in the same earnings quintile, working couple households % of working couples NOR (86-04) FIN (87-04) CZE (92-04) GBR (86-04) CAN (87-04) [AUT (94-04)] [HUN (94-05)] [FRA (84-00)] USA (86-04) NLD (83-04) ITA (87-04) [LUX (85-04)] ISR (86-05) 1 [ESP (90-04)] [GRC (95-04)] [MEX (84-04)] [POL (92-04)] AUS (85-03) DEU (84-04) [IRL (94-04)] [BEL (85-00)] 23 0 DNK (87-04) SWE (81-05) NOR (86-04) CAN (87-04) FIN (87-04) [FRA (84-00)] CZE (92-04) USA (86-04) [ITA (87-04)] GBR (86-04) [GRC (95-04)] [HUN (94-05)] NLD (83-04) [LUX (85-04)] [AUT (94-04)] ISR (86-05) 1 [MEX (84-04)] [ESP (90-04)] [POL (92-04)] AUS (85-03) IRL (94-04) [BEL (85-00)] DEU (84-04) Note: Refers to couple households with both partners working. Earnings refer to net earnings for countries in brackets and to gross earnings for other countries. Note: Refers 1. Information to couple households on data for Israel: with both partners working. Earnings refer to net earnings for countries in brackets and to gross earnings Source: for other Secretariat countries. calculations from the Luxembourg Income Study (LIS). 1. Information on data for Israel: Source: Secretariat calculations from the Luxembourg Income Study (LIS). The average degree of assortative mating, under this broader measure, increases from 34% to almost 40%

23 These data come from the data appendix to Atkinson, Piketty and Saez (2009), with additional information from country delegates in some cases. The information all Figure comes 9.1. from Toptax 1% records, incomeapart share, from the data for Finland. Figure 9.1. Top 1% income share, % 30 Australia New Zealand Canada United Kingdom Ireland United States Source: Alvaredo et al. (2011). Country delegate information: Australia ( ) and Canada ( ) Source: Income refers to individual income. Alvaredo et al. (2011). Country delegate information: Australia ( ) and Canada Figure ( ). 9.1 shows the shares of the top percentile group in pre-tax income for the English speaking countries from 1910 to 2008 (or the latest available Divided We Stand year). Extract The data show

24 Table 9.1. Share of top 1% in selected years Table 9.1. Share of top 1% in selected years III.9. TRENDS IN TOP INCOMES AND THEIR TAX POLICY IMPLICATION Australia Belgium Canada Denmark Finland France Germany Ireland Italy Japan Netherlands New Zealand Norway Portugal Spain Sweden Switzerland United Kingdom United States Note: The The data data in in the the first first four four columns relate relate to the to the nearest nearest available available year. year. Source: Alvaredo et al. et al. (2011). (2011). Country Country delegate delegate information: information: Australia Australia ( ), ( ), Canada ( ), Canada ( ), Italy Italy ( ), ( ), Netherlands Netherlands ( ), ( ), Norway Norway ( ) ( ) and Switzerland Switzerland ( ); ( ); Roine Roine and Waldenström and Waldenström (2008): Sweden (2008):( ). Sweden ( ). 1 2 Divided We Stand Extract