WHAT’S SO BAD ABOUT MORE INEQUALITY?

Lars OsbergEconomics DepartmentDalhousie University

ACADEMY OF THE SOCIAL SCIENCES IN AUSTRALIA – 2013 ANNUAL SYMPOSIUM Canberra, Australia November 12, 2013

“More Inequality” • Cross-national level comparisons @ point in time (e.g. U.S.t > Aus.t)

• Menu of choices? Causes ∆ health, happiness, crime, social mobility ?• Stability assumed – steady state Equal Growth rate @ all percentiles

• Over-time for same society – e.g. US2013 > US1983

• U.S., Australia, Canada – 30 years of Unbalanced Growth• Increasing Inequality Differential in growth rates: Top 1% >> Bottom 99%• Why expect big slowing of top 1% growth OR big acceleration of 99% growth?

• Continued differential in income growth rates plausible – compounds to ever larger gaps

• Income = Consume + Save: Implications of continued growth differentials?• Save: ↑ Financial Assets => ↑ Financial Liabilities =>↑ Debt Fragility=> Real Crises• Spend: => ↑ Extravagance; ↑ Advertising Luxuries; ↑ political & social advantages

• Increasing Inequality cannot be a steady state• Interacting Instabilities of Imbalances – but what next?

“More Inequality” – U.S.t > Canadat

Cross – national comparisons of levels• Reliable cross-national data on inequality only since 1970s

• Now a large literature on income measurement, equivalence scales, etc.

Socially important “Possibility Proof”

• Market Economies have widely varying levels of income inequality while competing successfully in global markets.

• i.e. There Are Alternatives – different choices in different places

A menu of choices?

Icelan

d

Sloven

ia

Norway

Denmark

Czech

Rep

ublic

Finlan

d

Slovak

Rep

ublic

Belgium

Austria

Sweden

Luxe

mbourg

German

y

Netherl

ands

France

Poland

Korea

Estonia Ita

ly

Canad

a

Austra

lia

Greece

Spain

United

King

dom

Portug

alIsr

ael

United

Stat

es

Mexico

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Income Inequality in OECD - 2010Gini Index of Equivalent Disposable Income

What would more equality imply ?• More Equality causes more

• health• life expectancy• trust• social mobility• educational performance AND LESS• infant mortality• Violence• obesity• mental illness• teen births• homicides• Imprisonment

• Wilkinson & Pickett: The Spirit Level: Why Equality is Better for Everyone (+ many articles)

• Method: Cross-national correlations & scatterplots, primate & workplace studies

• H0: ↑inequality =>↑ stress of social interaction

• Is Inequality Guilty of all this?• Can Inequality be proved Guilty?

• Onus of proof ?• 95% Prob (harmless) OR 95% Prob (Harmful)

• Level of certainty ?• “Balance of Probabilities” or “Beyond Any Doubt”

“too many theories for the number of available data points” - Leigh et al (2009:399)

• Multiple Plausible Indicators of Complex Concepts• e.g. “Health” & “Inequality”; => ambiguity of estimates

• Causation – very hard to prove: formal econometrics not feasible• Outliers – weird or very informative ?• Onus of proof – required proof: “harmful” or “harmless” ?

• Most Convincing evidence:• Intergenerational Social Mobility & Inequality of Opportunity

• Also – more inequality => more unhappiness & social conflict

Robust association – inequality of outcome & inequality of opportunityBrunori, Ferreira, Peragrine (2013:27)

Variety of Social Mobility Measures

Intergenerational - Correlation Education - Earnings elasticity - Decile transitions

- All are lower where inequality of income is greater

Equality of Opportunity?

• Parents choose Human Capital Investment for own Children subject to Income Constraint

• Becker/Tomes: parental altruism model • Max U0 = u0(C0,u1(C1, U2))

• s.t. Yi = Ci + HKBi + Ki • Yi = Wi + rhi HKBi-1+ rk Ki-1.

Market Society Implies:• Inequality of Outcome in one generation begets Inequality

of Opportunity in next generation

• Pure Market Economy is Dynastic Society • (subject to random variation in rhi and rk)

Not a new insight – Marshall & many others

Assumed: No scarcity of top slots• Human Capital Model assumes no rationing of access to top slots

• Harvard admits all applicants who can pay; All hard-working MBAs become CEO

• Strong Assumption: There is nothing competitive about life.• success by others does not affect probability (success by self)

• BUT in a competitive race, only top few can win• Scarcity of top slots => own prob (success) decreases when others prob (success) increases• Intergenerational mobility in social rank: trading ranks - when some go up, others must go down

• Implications of Rationing of Access + Increased Payoff to top slots?• “rat-race” model → greater over-investment in effort to increase own Prob (promotion)

• Increasing stakes in early school success imply more pressurized childhood ?• Real “Equality of Opportunity” has greater costs to affluent parents

• Greater “drop from top” implies less support by affluent for public spending to equalize opportunity (which would decrease chances of own kids’ success)

Cross-National Comparisons – Stability of Inequality level is assumed

• Steady State Inequality Equal Growth rate @ all percentiles• Happy Accident of 1953 -1980• Not our current problem

• “More Inequality” - U.S. & Canada & Australia• Increasing Inequality over time Unbalanced Growth by Income class

• Increasing inequality cannot be a steady state• Unbalanced Growth => Interacting Instabilities of Imbalances

No stable level of Gini Index

Canada- Rising – esp. since 1990s- 2000+

- ↓ middle offset ↑ top- Top-coding survey data

USA Rising since early 1980s

Australia Trending up 1995 2000 2004 2008 2010

0.25

0.27

0.29

0.31

0.33

0.35

0.37

0.39

Australia, Canada, USA & OECDGini Index of Post-Tax/Transfer Equivalent

Household Income

Australia CanadaUnited States ALL OECD

Alvaredo, Atkinson, Piketty, Saez (2013)

“most of the action has been at the very top”

U.S. & Canada – lower percentiles show little change in real income 1980 -2012

Australia: resource boom => ↑ earnings => change in bottom 99%

Income Share = Ratio

• Income Share of Top 1% = Incomes of Top 1% Incomes of 99% + Incomes top 1%

• So where has the action been in Income Shares? • Numerator (Real Incomes of top 1%) ?• Denominator (Real Incomes of Bottom 99%) ?

Increasingly higher long-run growth rates at top

U.S.& Canada – little growth in bottom deciles

Australia – significant earnings growth for 90%

Top /Bottom DifferentialIn income growth rates was similar

Focus on Top 1% - approximation to ↑ growth rate

Bottom

90% av

erage

inco

me

Top 10

-5% av

erage

inco

me

Top 5-

1% av

erage

inco

me

Top 1-

0.5%

avera

ge in

come

Top 0.

5-0.1%

avera

ge in

come

Top 0.

1-0.01

% avera

ge in

come-0.50%

0.00%0.50%1.00%1.50%2.00%2.50%3.00%3.50%4.00%4.50%

Average Real Income Compound Annual Growth Rate: 1982-2010

AUSTRALIA, CANADA, USA

USA CANADA AUSTRALIA

Top 1% Income - No Natural Upper Bound

Real Average Income Top 1% - Cyclical Fluctuations - Upward trend -slow 1935-1980 - accelerates 1985+

CCPC income not included in Canadian data

1913

1918

1923

1928

1933

1938

1943

1948

1953

1958

1963

1968

1973

1978

1983

1988

1993

1998

2003

2008

0

200000

400000

600000

800000

1000000

1200000

TOP 1% AVERAGE REAL INCOME Australia, Canada & USA

USA CANADA AUSTRALIA

U.S.Balanced Growth = atypical episode

- 1965-1980 - equal growth rates for

top 1% & bottom 99%

- - 1940 – 1964- higher growth rates at

bottom – especially 1940s

- - 1980 +- Much higher growth

rates for top 1% -0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

0.1

0.12

REAL INCOME GROWTH RATES: USATOP 1%, BOTTOM 99% & 90%

10 YEAR COMPOUND ANNUAL RATE

TOP 1% BOTTOM 99% BOTTOM 90%

Canada:Longer balanced growth period Mid 1950s-mid 1980s: - bottom 90% growth rate

slightly higher than top 1% (but roughly balanced)

Pre-1950s & post 1985:- Significant differences in

income growth rates

- Pre-1950 – compression- Post 1986 – top-end

growth much faster

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

-0.04

-0.02

-6.93889390390723E-18

0.02

0.04

0.06

0.08

Figure 7REAL INCOME GROWTH RATES: CANADA

TOP 1%, BOTTOM 90%10 YEAR COMPOUND ANNUAL RATE

TOP 1% BOTTOM 90%

“Once-only” & Income Growth 1940-1970• Recovery from Mass Unemployment of Depression + WWII controls• Structural Changes with Major Income Impacts

1. High % agriculture => rural out-migration => big wage gains2. Low % employed women => big impact of increase female jobs 3. Low % complete post-secondary => high marginal HK returns4. Capital deepening => increased MPL post WWII 5. “Baby Boom” => demographic bulge 6. Unionization; increased bargaining power until late 1950s

• Political economy of social policy ?• Credible ‘hard left’ political option => “threat effect” for elites

AustraliaUnequal growth – normal eventNot same pattern as U.S. & Canada pre 1980s

35 years of compression 1951-1986

1986 + similar differential in growth rates

-0.08

-0.06

-0.04

-0.02

-1.38777878078145E-17

0.02

0.04

0.06

0.08

0.1

REAL INCOME GROWTH RATES: AUSTRALIA TOP 1%, BOTTOM 99% & 90%

10 YEAR COMPOUND ANNUAL RATE

TOP 1% BOTTOM 99% BOTTOM 90%

Suppose Past U.S. Trends Continue ?

Median Top 1% Dollar AnnualTop 1%/Median

Household Average Gap $ increase Ratio

1984

47,181 376,135

328,954

10,341 8.0

2011

51,100 878,960

827,860

26,025 17.2

2031

54,207

1,632,378

1,578,172

49,613 30.1

2038

55,338

2,027,304

1,971,966

61,992 36.6

Growth Rate 0.30% 3.14%

1984-2011

- 1984-2011: Annual growth rate differential ~ 2.84%

- No Big Deal if 2-3 years

- Compounds to very large $ differentials & ratios over 20+ years

- Too Large to Believe?

- Why would income growth rates change?

Framing the question ? • Increasing Level of Income Inequality ? OR

• Differential in long term income growth rates ?• Top 1% income growth rate (3.14%) >> Bottom 99% growth rate (0.3%)

• Different words for same reality BUT• Differential Growth Rates perspective suggests:

• Why did growth rates differ ?• Why would growth rates equalize ?

• Substantial Slowing of Top 1% ?• Big Acceleration of 99% ?

• One-time level changes cannot explain long-term trend differentials• E.g. need series of tax cuts & continual ↑ labour supply

1984

1989

1994

1999

2004

2009

2014

2019

2024

2029

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

Top 1% to Median Income Ratio USA

actual2.84% differential3.36% differential

Auto-equilibrating Market Mechanisms ?• Top 1% Income: Not a Capital / Labour Factor Shares story

• Large % income of top 1% = Labour compensation

• Why might top 1% growth slow?• Labour Market Story needed – could it be ↑ Supply (Effort) <= Tax cuts ?

• Could Top 1% run out of steam ? (i.e. @ max. possible effort) ?• “Effort” = (Hours per year)*(Work Intensity per hour)

• Max (Annual Hours) = 6,000 ?? (16*365=5,840) ; Intensity has some upper bound• BUT were the elite of 1982 really that slack ? [top 0.1% 1982 = 0.326 top 0.1%2011)]

+ timing does not fit + Labour/leisure choice is levels model & => backward-bending SSL

H0: Segmented Labour Markets ?• “Globals and their peers”

• Top corporate teams share in monopolistically competitive profits• Rents to hierarchical rank increase with rank

• Profits = f (firm size <= scale of market) • Post 1980 – ↓trade barriers, ↑ firm growth rate <= global market growth; • Sets benchmarks for top positions in national firms & non-profit sector

• U.S. leads Anglo wage contours, with slow filter to other national top ends

• “Locals”• Long run growth rate hourly wage ≤ labour productivity growth

+ Share of Resource sector rents if unions or rapid development; - ∆ wage <= slack labour markets (if Ut > U*)

• Implication: Differential in Income growth rates persists

What plausible alternative model implies likely:- substantial slowing of top 1% or- big acceleration of rest ?

• Could more education sufficiently accelerate the long-run growth rate of average 99% income?• U.S., Canada, Australia – already well educated

• Diminishing returns at successively smaller margin• Equalization within 99% does not imply acceleration of average 99%• Educational reform – inherently long lags to pay off• 25-64 Tertiary Education : 51% Canada > 42% U.S. > 38% Australia

• No evidence of convergent middle class incomes in Canada

Stable Inequality Balanced Growth IFF Same Rate Income Increase @ all income percentiles

• BUT U.S.: Annual Income growth rate 1984-2011: • Top 1 % = 3.14% = r1 ; Median household = 0.3 % = rm

• Income Levels diverging @ r1 - rm = 2.8%

• Short-run chances for rm = 3%?• Unions weak; Low-wage competition strong; slack labour demand

• Why would Income Setting @ Top change & long run r1 ↓ ?• What are implications of continued Unbalanced Growth ?

Income = Savings + Consumption• Income Increases @ top => Increase Savings => Increase Loanable Funds

• Macro Real Expenditure Balance requires: Increased Savings top 1% = increased spending rest

• Save => purchase of financial asset• Financial Assets = Financial Liabilities

• Financial Instrument: Asset for Holder = Liability for Issuer

Differential Flows accumulate to Stocks

DEBATE : ↑ inequality of consumption < ↑ inequality of income ? If true: mitigates short run utility implications of greater inequality

• IGNORED: If true: implies changing distribution of assets and liabilities

• ↑ Net Savings @ top imply Accumulating Debts @ bottom• Savings & debts grow @ r1 but median income grows @ rm => ↑ leverage

• Financial Fragility => Financial Crises => Real Recessions (Kumhof & Ranciere)

• Recessions => Counter-cyclical stimulus => ↑ Public Debt / GDP => unpleasant choices for continued monetization or austerity / contraction

Debt StabilityDt = (1 + rt)* Dt-1 - PBt

Dt = Debt in period trt = average rate of interest in period tPBt = Primary Balance in period t

= (Receiptst – Expenditurest)

∆ (D/Y)t = (rt - gt)*(Dt-1/Yt ) - (PBt / Yt) Yt = GDP for nation; Household Income for familiesgt = growth rate∆ (D/Y)t = change in Debt/Income ratioWill rt < gt forever?

Debt Instability – not just a Public Sector Problem !

∆ (D/Y)t = (rt - gt)*(Dt-1/Yt ) - (PBt / Yt)

• Debt overhang compounds if / when: rt > gt

• Accumulated Deficits => ↑ Debt/GDP => Deficit => ↑ Debt => etc.• Can & Aus: low rt to maintain demand enables ↑ household leverage

• Unpalatable Choices:• Anti-Inflation Monetary Policy increases (rt - gt) at both ends• Can rt < gt for long-term ? How to unwind rising household leverage?

Instability <= Increasing Consumption?• Macro: Extravagant Elite Consumption does recycle Income

• “Downton Abbey” – high & stable inequality • Norms of consumption & deference built up over many decades

• In time, habituation → ”natural order of things” for both servants & served

• Increasing Income gaps imply Increasingly Extravagant Elite Consumption required for Macro balance• norms of luxury → increasingly distant from median• Veblen: “conspicuous consumption” = the main point of great wealth

• “if you’ve got it, flaunt it” lifestyles are resented by some• Gaps Increasing over time @ r1 - rm

The Increasing Advertising of Envy• Increasing top 1% share = Increasing market for luxury goods

• Increasing % of advertising for luxury / status goods

• Status goods – a pointless purchase if nobody else thinks it’s “special / desirable / exclusive”

• => Advertising must increasingly emphasize exclusivity / luxury / privilege• Increasingly remind 99% of what only 1% can afford (& 99.9 of 0.1% )

• Increasing Inequality increases Market Incentives to manufacture envy.

Externalities of top 1% spending ?• Increasingly distant top incomes imply:

• Increasing market for infrastructure of exclusivity• Separate world of resorts, gated communities, restaurants, etc.

• Increasingly difficult to socialize across income classes

• Escalating Consumption Norms? – set @ top & ripple down ? (Frank)

• => Increased middle class debts & increased financial fragility

• BUT: Why not just ignore (& tax) the top 1% ?

Externality 1 – political influence• Top 1% refuse to be ignored politically

• U.S. evidence:• political & social preferences of top 1% quite different from 99%• Top 1% much more active politically than the 99%• campaign funding depends heavily on major donors • legislation heavily influenced by the policy priorities of top 1%

• Political influence: More for 1% implies less for 99%

• “Deeper Pockets” & Meaningful Democracy ?

Externality 2: Advantages for kids• Increasingly affluent families buy increasingly more advantages for their

children

• “Income effect” of rising real incomes (Normal good) PLUS

• “Price effect” – Increasing “drop from top” implies ever greater incentives to prevent downward social mobility for own children• Top 1% / Median ratio increasing over time => ↑ cost of move from top to median

• When top 1% avoid downward mobility of their own kids, decreases the chances of upward mobility for 99%

• Maintaining belief in “equality of opportunity” becomes ever harder

If markets will not auto-equilibrate, can political economy stabilize ?USA: What chance for a New “New Deal” ?

• 1930s: FDR & “New Deal”• U.S. Policy Innovation Stabilized Growth & Inequality

• Cyclical Stimulus + Structural Reforms + Progressive Taxation + Social Security • Restraint top end income growth + fiscal recycling + financial market

regulation + unions => ↓ inequality + period of balanced growth

• “Tax & Spend” can in principle stabilize the distribution of post-fisc income for any given trend in market income. – but how likely is that?

The unsustainable does not last – but what follows?

• Unbalanced Income Growth Ever Increasing Inequality • Cannot be a steady state equilibrium

• Produces Interacting Instabilities – with cumulative impacts

• Parallels with 1930s but many structural changes since

• No Automatic Economic self-correction Tendency is apparent• Political Economy of Adaptation to Systemic Instability:

• Europe in 1930s: both disastrous choices and enduring successes• Political choices matter

Comments very much appreciated:

lars.osberg@dal.ca

Differences in Rates of Growth Drive Changing Income Shares1940-1973 – strong growth in bottom 99% incomes + slow growth for top 1% = declining share for top 1%

1980-2012 – income stagnancy for bottom 99% + strong growth for top 1% = rising income share for top 1%

T. Piketty and E. Saez “Income and Wage Inequality in the United Staes, 1913-2002,” Chapter 5 in The Oxford Handbook of Economic Inequality, edited by Wiemer Salverda, Brian Nolan, and Tim Smeeding, Oxford University Press, Oxford, 2009, page 174

Canada – nil real growth for most

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

0

20000

40000

60000

80000

100000

120000

Total Income of Canadian Family Units:1976-2009

20th percentile 40th percentile median 60th percentile 80th percentile

2009

Dol

lars

U.S. – real growth only at top

USA: Conflicted attitudes + $ politics• Bimodal distribution → small

migration tips majority balance• BUT short terms + division

powers + courts => gridlock + soon tips back

• “Deeper Pockets” • Increased economic Inequality =>

Increased Inequality of Political Influence

-0.50

0

-0.37

5

-0.25

0

-0.12

50.0

000.1

250.2

500.3

750.5

000.6

250.7

500.8

751.0

001.1

251.2

501.3

751.5

001.6

251.7

501.8

752.0

000

0.2

0.4

0.6

0.8

1

1.2

1.4

Preferences for Leveling in the United States 1987-1999

U...

Rel

ativ

e fre

quen

cy

1925

1928

1931

1934

1937

1940

1943

1946

1949

1952

1955

1958

1961

1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

2009

-0.080

-0.060

-0.040

-0.020

0.000

0.020

0.040

0.060

0.080REAL INCOME GROWTH RATES: FRANCE

TOP 1%, BOTTOM 90%10 YEAR COMPOUND ANNUAL RATE

TOP 1% BOTTOM 90%

60

30

10

A party that promised to raisetaxes on the richA party that promised not toraise taxesDK/NR

BASE: Canadians; February 21-28, 2012 (n=3,699)

Q. In the next federal election, would you be more likely to support a party that promised to NOT raise taxes or a party that promised to raise taxes on the rich?