Is Inequality Increasing?

Presentation for Parliamentary Library Vital Issues Seminar, 10 October 2012Peter Whiteford, Crawford School of Public Policypeter.whiteford@anu.edu.au

Footer text goes in here

Background and outlineThis presentation is primarily descriptive what has happened to income inequality in Australia over recent decades, and what factors appear to be associated with these trends. Data and methods; measures of inequalitySituating Australia internationallyIncome inequality: trends and driving forcesRedistribution through taxes and benefitsDiscussion and conclusions

Data and methodsData are from ABS income surveys from 1981-82 to 2007-08 and in some cases 2009-10. The ABS has changed and improved income measures over time; for consistency we use the unimproved income measure, showing lower inequality after 2005-06, but effects on earlier trends uncertain.Income measure is current weekly income of income units (nuclear family), adjusted for household size using revised OECD equivalence scales.Income is made up of market income (earnings, self-employment, investment and property income, private transfers); the addition of transfers from government (social security benefits) or privately (e.g. child support) produces gross income; direct taxes are deducted to estimate cash disposable income.Interpreting changes as result of government policy decisions is problematic e.g. unemployment rose rapidly between time of 1981-82 Income Survey and election of Labor government in March 1983; declines in welfare receipt after 2000 partly reflect 1995 reforms (raising pension age for women, phasing-out dependency payments).There are also long-term cohort effects - e.g. rising educational attainment of women and increase in female labour force participation; declines then increases in employment of older workers.Some important policy changes not fully captured in cash disposable incomes e.g. reintroduction of Medicare, extension of superannuation, introduction of GST. However, policy trade-offs accompanying these changes may be incorporated e.g. wage restraint under Accord.

How is inequality measured?What inequalities are we interested in or concerned about inequality of income, inequality of opportunity, inequality of wealth?Inequality is often thought of as disparities or gaps what is the distance between a low income and a high income household, or what is the ratio of their incomes, e.g. what is the gap between the richest 10% and the poorest 10% (decile) or the richest and poorest 20% (quintile)?A lot of inequality is at the extremes, i.e. within the richest and poorest income groupsThe measure of inequality most commonly used is the Gini coefficient, which varies between zero when all households have exactly the same income and one when one household has all the income. Preferred because it is calculated for everyone in the population.In 2003 the Gini coefficient for Australia was 0.301 and the 90/10 ratio was around 4 to 1; in the USA, the Gini was 0.381 and the 90/10 ratio was about 6 to 1; in Denmark the Gini was 0.232 and the 90/10 ratio was 2.7 to 1. The ratios of the average incomes of the top decile to the average incomes of the bottom decile were 4.6 to 1 (Denmark), 7.2 to 1 (Australia) and 16 to 1 (USA).

Level of inequality in OECD countries20052008

Chart1

0.22DenmarkDenmark

0.23SwedenSweden

0.26LuxembourgLuxembourg

0.26AustriaAustria

0.26Czech RepublicCzech Republic

0.27SwitzerlandSwitzerland

0.27Slovak RepublicSlovak Republic

0.27FinlandFinland

0.27NetherlandsNetherlands

0.27BelgiumBelgium

0.27FranceFrance

0.28IcelandIceland

0.28NorwayNorway

0.29HungaryHungary

0.3GermanyGermany

Australia0.3Australia

0.31OECDOECD

0.31KoreaKorea

0.31JapanJapan

0.32PolandPoland

0.32CanadaCanada

0.32SpainSpain

0.32GreeceGreece

0.33IrelandIreland

0.34New ZealandNew Zealand

0.34United KingdomUnited Kingdom

0.37ItalyItaly

0.38USAUSA

0.42PortugalPortugal

0.43TurkeyTurkey

0.47MexicoMexico

Column3

Column1

Series 3

Sheet1

Column3Column1Series 3

Denmark0.222

Sweden0.232

Luxembourg0.263

Austria0.265

Czech Republic0.26

Switzerland0.27

Slovak Republic0.27

Finland0.27

Netherlands0.27

Belgium0.27

France0.27

Iceland0.28

Norway0.28

Hungary0.29

Germany0.3

Australia0.3

OECD0.31

Korea0.31

Japan0.31

Poland0.32

Canada0.32

Spain0.32

Greece0.32

Ireland0.33

New Zealand0.34

United Kingdom0.34

Italy0.37

USA0.38

Portugal0.42

Turkey0.43

Mexico0.47

To resize chart data range, drag lower right corner of range.

Chart1

0.23DenmarkDenmark

0.23SwedenSweden

0.27AustriaAustria

0.27BelgiumBelgium

0.27Czech RepublicCzech Republic

0.27FinlandFinland

0.27NetherlandsNetherlands

0.28FranceFrance

0.28IcelandIceland

0.28NorwayNorway

0.28SwitzerlandSwitzerland

0.29HungaryHungary

0.3GermanyGermany

0.31OECDOECD

0.32CanadaCanada

0.32GreeceGreece

0.32JapanJapan

0.32SpainSpain

Australia0.328Australia

0.33IrelandIreland

0.34New ZealandNew Zealand

0.34United KingdomUnited Kingdom

0.35ItalyItaly

0.38PortugalPortugal

0.38United StatesUnited States

0.43TurkeyTurkey

0.47MexicoMexico

Series 1

Series 2

Series 3

Sheet1

Series 1Series 2Series 3

Denmark0.232

Sweden0.232

Austria0.273

Belgium0.275

Czech Republic0.27

Finland0.27

Netherlands0.27

France0.28

Iceland0.28

Norway0.28

Switzerland0.28

Hungary0.29

Germany0.3

OECD0.31

Canada0.32

Greece0.32

Japan0.32

Spain0.32

Australia0.328

Ireland0.33

New Zealand0.34

United Kingdom0.34

Italy0.35

Portugal0.38

United States0.38

Turkey0.43

Mexico0.47

To resize chart data range, drag lower right corner of range.

Change in inequality, OECD countries, 1995 to 2007-08

Chart1

-0.081TurkeyTurkey

-0.043MexicoMexico

-0.033ChileChile

-0.031IrelandIreland

-0.029GreeceGreece

-0.028BelgiumBelgium

-0.026SpainSpain

-0.022HungaryHungary

-0.011ItalyItaly

-0.006PortugalPortugal

-0.005New ZealandNew Zealand

-0.003NetherlandsNetherlands

-0.001Czech RepublicCzech Republic

0.006JapanJapan

0.006United KingdomUnited Kingdom

0.007NorwayNorway

0.016FranceFrance

0.017United StatesUnited States

0.023AustriaAustria

0.027AustraliaAustralia

0.029GermanyGermany

0.029LuxembourgLuxembourg

0.033IsraelIsrael

0.033DenmarkDenmark

0.035CanadaCanada

0.041FinlandFinland

0.048SwedenSweden

Series 1

Column1

Column2

Sheet1

Series 1Column1Column2

Turkey-0.081

Mexico-0.043

Chile-0.033

Ireland-0.031

Greece-0.029

Belgium-0.028

Spain-0.026

Hungary-0.022

Italy-0.011

Portugal-0.006

New Zealand-0.005

Netherlands-0.003

Czech Republic-0.001

Japan0.006

United Kingdom0.006

Norway0.007

France0.016

United States0.017

Austria0.023

Australia0.027

Germany0.029

Luxembourg0.029

Israel0.033

Denmark0.033

Canada0.035

Finland0.041

Sweden0.048

To resize chart data range, drag lower right corner of range.

Change in Sen welfare index, OECD countries, 1995 to 2008Change in real mean household income adjusted for inequality

Change in real GDP, Australia and selected OECD countries, 2008 to 2011Q4 2007=100

Median Australian households have fared extremely well 1998=1AustraliaUSA

Chart1

11

1.04017857141.022

1.06696428571.044397463

1.07589285711.067653277

1.10267857141.0821

1.12276785711.0951374207

1.156251.1881606765

1.17410714291.24

1.19196428571.2832980973

1.218751.37

1.24330357141.4608879493

1.23214285712009

Real GDP per capita

Real median household income

Sheet1

199819992000200120022003200420052006200720082009

Real GDP per capita11.04017857141.06696428571.07589285711.10267857141.12276785711.156251.17410714291.19196428571.218751.24330357141.2321428571

Real median household income11.0221.0443974631.0676532771.08211.09513742071.18816067651.241.28329809731.371.4608879493

To resize chart data range, drag lower right corner of range.

Chart1

11.0000031709

1.03008720591.0242940046

1.0487976871.0291466173

1.04364995421.0145854119

1.05232353151.0039011066

1.06760219071.000289475

1.09724278961.0004342576

1.12018428411.0062115342

1.13908281031.0119184905

1.15088263641.0339316386

1.12213525140.9979041356

1.09731330650.9786220766

1.11614131580.9678841007

Real GDP per capita

Real median income

Sheet1

Column1Real GDP per capitaReal median income

199811.0000

19991.03008720591.0243

20001.0487976871.0291

20011.04364995421.0146

20021.05232353151.0039

20031.06760219071.0003

20041.09724278961.0004

20051.12018428411.0062

20061.13908281031.0119

20071.15088263641.0339

20081.12213525140.9979

20091.09731330650.9786

20101.11614131580.9679

Trends in income inequality in Australia, 1981-82 to 2009-10 Gini coefficient

Trends in income inequality (Gini coefficient) among households with a head aged 65 years and over, Australia, 2000-2001 to 2009-10

Chart1

0.2360.231

0.2960.262

0.2840.289

0.3090.269

0.3390.298

0.3320.259

Couples 65 and over

Singles 65 and over

Sheet1

Couples 65 and overSingles 65 and overCouples 65 and overSingles 65 and over

2000-20010.2360.2310.6844349680.5842217484

2002-030.2960.262

2003-040.2840.289

2005-060.3090.269

2007-080.3390.2980.68803945750.5351418002

2009-100.3320.2590.70047169810.5577830189

Patterns of income growth by decile and period, Australia, 1981-82 to 2007-08 Average annual percentage change in real equivalent income unit income, working age

Chart1

0.02204011940.01508991820.0463779926

0.00976149170.01832189960.066688636

0.00285826650.02452529510.0697716637

0.00303800290.02607744810.0629399586

0.00494275060.02447778850.0580107157

0.00576918010.02506260910.0553264466

0.00621536850.02405766280.0556420784

0.00704278890.02330404980.0538232088

0.00662641180.02521465060.0582011804

0.0116118020.0277421540.0940063124

1982 to 1996

1996 to 2003

2003 to 2007

Sheet1

1982 to 19961996 to 20032003 to 20072007 to 2009

12.2%1.5%4.6%-4.6%50.8%

21.0%1.8%6.7%-3.7%45.7%

30.3%2.5%7.0%-5.4%38.2%

40.3%2.6%6.3%-3.2%41.3%

50.5%2.4%5.8%-1.7%43.8%

60.6%2.5%5.5%-1.6%44.5%

70.6%2.4%5.6%-0.8%46.2%

80.7%2.3%5.4%0.6%48.9%

90.7%2.5%5.8%1.2%52.7%

101.2%2.8%9.4%-4.7%64.0%

To resize chart data range, drag lower right corner of range.

Trends in income inequality in different income components among working age income units, Australia, 1982 to 2007-08Concentration coefficient

Earnings have grown with rising employmentAnnual average real change in earnings by deciles of income unit incomeWomenMen*

Trends in income inequality in different income measures among working age income units, Australia, 1982 to 2007-08

Reduction in inequality among income units of working age, Australia, 1982 to 2007-08Point difference in Gini coefficient

Change in real value of transfers (2008 $pw) received by deciles of working age income units*

Change in working age income support recipients, 1996-97 to 2009-10% of households by age group

Joblessness is a major source of inequality in AustraliaEarnings represent around three-quarters of total pre-tax household income, the largest single component.In 1983, a full-time worker at the 90th percentile earned 2.0 times as much as a worker at the 10th percentile- this disparity increased to 2.3 in 1996, 2.5 in 2004, and 2.8 in 2009-10.In 1982 a working-age family at the 90th percentile earned 112 times as much as a family at the 10th percentile this disparity reduced to 56 times as much in 1996 and 49 times as much in 2009-10.When social security benefits are added in, this disparity was reduced to 9.8 times as much in 1982, 10.2 times as much in 1996 and 11.4 times as much in 2009-10.Taxes reduced the disparity to 8.0 to 1 in 1982, 7.5 to 1 in 1996 and 9.6 to 1 in 2009-10.

Poverty in a time of prosperityPayments for single person as % of median equivalent income

Chart1

0.46632640030.5030874119

0.46948510720.5065269258

0.46130328580.4977089157

0.45651272630.5114840036

0.43632534090.4984129905

0.43938109950.5028321357

0.42645156270.5140478397

0.38206517220.478702874

0.35736677120.5251535232

Newstart

Pension

Sheet1

Column1NewstartPensionSeries 3

199447%50%2

199547%51%2

199646%50%3

200046%51%5

200244%50%

200344%50%

200543%51%

200738%48%

200936%53%

To resize chart data range, drag lower right corner of range.

Average tax rates (%) by deciles of household income, 1982, 1996-97 and 2007-08

Income measures and concepts are importantInequality is lower the longer the time period over which it is measured:Between 2001 and 2009 the Gini coefficient ranged between 0.306 in 2000, 0.300 in 2003, 0.312 in 2007 and 2008 and 0.299 in 2009; inequality over a two-year period varied between 0.286 and 0.295; inequality over a four year period varied between 0.273 and 0.282; inequality over the 9 years was 0.263. Before 2009 longer term measures showed increasing inequality.Broader measures of resources also suggest lower inequality: Taking account of non-cash benefits (health, education, community services) even after subtracting indirect taxes reduces inequality. For example, the 90/10 ratio for disposable income in 2009-10 was 3.94 to 1, but for final income it was 2.64 to 1 but up from 2.48 to 1 in 2003-04.Adding imputed income from housing reduces the Gini coefficient in 2009-10 from 0.328 to 0.309 but still shows increasing inequality (0.277 in 2003-04 and 0.290 in 2005-06).Inequality is mainly reduced by raising the bottom not reducing the top:Direct taxes and cash transfers increase the incomes of the 10th percentile by 231% and lower the income of the 90th percentile by 19%;The net effect of Indirect benefits and taxes is to raise the 10th percentile by a further 50% and raise the 90th percentile by 1%.*

Assessing income inequality trendsTrends in inequality differ by time period, income components and income measures. Thus, there is no single trend, but the complex interaction of multiple influences.Market income inequality rose in period of Labor government, mainly reflecting higher joblessness and wider wage dispersion. Disposable income inequality rose significantly less (about 1/3 as great), reflecting both transfer and tax changes. Real market incomes fell for the second, third and fourth deciles, but these declines were offset by increases in social security benefits.Despite increasing wage dispersion, market income inequality fell from 1996-97 to 2007-08, mainly because of increased family earnings, particularly for women. Capital income inequality rose significantly after 2003, but insufficient to offset lower inequality in earnings.The effectiveness of the tax system in reducing inequality was stable in the 1980s and early 1990s, but reduced after 1996. Average tax rates fell most for the highest decile (about 4 percentage points).The effectiveness of the transfer system in reducing inequality increased by about 40% in the period of the Labor government, but fell back to its original level by 2007-08. This does not necessarily reflect explicit policy change, but rising earnings among lower income groups lead to a scaling back of income support, heightened by indexation of key benefits to prices.

What are the drivers of inequality in Australia?Wage inequality has increased steadily from early 1980s onwards.Trends differ significantly by time period in the early 1980s and again in the early 1990s median income growth was very slow and there was a hollowing out of the middle class gains being highest at the top and bottom of the income distribution. Much of the increase in inequality was offset by taxes and transfers and more so if account is taken of non-cash benefits and indirect taxes.From the mid 1990s to the great recession income growth was very high by historic and international standards Australia had the highest income growth at the median of any country apart from Ireland. All income groups had large real income increases, but the richest did best. Taxes and transfers reduced inequality less effectively than in the mid 1990s. Even though market income inequality fell, disposable income inequality rose.After 2008, incomes fell somewhat and inequality fell, mainly due to large declines in incomes from property and investments at the top of the income distribution. The various household stimulus packages were very progressive, as was the large increase in age pensions

*

3. Occupy movement and 1%.*the new income measures now include non-cash benefits, bonuses, termination payments and payments for irregular overtime.Commencing with SIH 200708, interest paid on money borrowed to purchase shares orunits in trusts has been netted off income earned from these sources when deriving netincome estimates, with a similar revision applied to results for 200304 and 200506.While regular workers'compensation receipts have been included in previously published results, lump sumreceipts were not. Commencing in SIH 200708, both forms of workers' compensationare included in the published estimates.In SIH 200708, a wider range of data on financial support received from and paid tofamily members resident outside the household was collected. Previously these weremainly limited to regular payments for spousal maintenance and child support. In200708, respondents were asked to include other forms of financial support, includinggoods and services received which were purchased by others e.g. rent, education, food,clothing, car registration and utilities.implementation of the broader measure of income in 200708resulted in an $85 increase in mean weekly gross household income, compared to theprevious definition. The additional inclusions affected 3.4m households in total (43%).Most of the impact was on employment income, which increased by $89 per week onaverage. The inclusion of non-cash employment benefits and bonuses had the mostimpact ($43 and $32 per week respectively).worked.***