introduction to taxes and transfers: income distribution ...saez/course131/taxintro_ch17.pdf ·...

Introduction to Taxes and Transfers:

Income Distribution, Poverty, Taxes and

Transfers

131 Undergraduate Public Economics

Emmanuel Saez

UC Berkeley

1

REMINDER: Two General Rules for Government

Intervention

1) Failure of 1st Welfare Theorem: Government interven-

tion can help if there are market failures

2) Fallacy of the 2nd Welfare Theorem: Distortionary

Government intervention is required to reduce economic in-

equality

2

Role 2: 2nd Welfare Theorem Fallacy

Even with no market failures, free market outcome might gen-erate substantial inequality

Inequality matters because people evaluate their economicwell-being relative to others, not in absolute terms ⇒ Pub-lic cares about inequality

2nd Welfare Theorem: Any Pareto Efficient outcome canbe reached by (1) Suitable redistribution of initial endowments[individualized lump-sum taxes based on indiv. characteristicsand not behavior], (2) Then letting markets work freely

⇒ No conflict between efficiency and equity

In reality, redistribution of initial endowments is not feasible(information pb) ⇒ govt needs to use taxes and transfersbased on economic outcomes (income, consumption, etc.) ⇒Conflict between efficiency and equity

3

Income Inequality: Labor vs. Capital Income

Individuals derive market income (before tax) from labor and

capital: z = wl + rk where w is wage, l is labor supply, k is

capital, r is rate of return on capital

1) Labor income inequality is due to differences in working

abilities (education, talent, physical ability, etc.), work effort

(hours of work, effort on the job, etc.), and luck (labor effort

might succeed or not)

2) Capital income inequality is due to differences in wealth

k (due to past saving behavior and inheritances received), and

in rates of return r (varies dramatically overtime and across

assets)

4

Income Inequality: Labor vs. Capital Income

1) Capital Income (or wealth) is more concentrated than Labor

Income: Top 1% wealth holders have 1/3 of total wealth.

Top 1% labor income earners have about 15% of total labor

income. [Top 1% incomes have 20% of total income]

2) Labor income is around 80% of aggregate market income

from National Accounts (capital income is 20%). Fairly con-

stant overtime and across industrialized countries.

[In GDP, gross capital share is higher (30%) because it includes

depreciation of capital]

5

Income Inequality Measurement

Inequality can be measured by indexes such as Gini, log-variance,

quantile income shares which are functions of the income dis-

tribution F (z)

Gini = 2 * area between 45 degree line and Lorenz curve

Lorenz curve L(p) at percentile p is fraction of total income

earned by individuals below percentile p

0 ≤ L(p) ≤ p

Gini=0 means perfect equality

Gini=1 means complete inequality (top person has all the in-

come)

6

Gini Coefficient California pre-tax income, 2000, Gini=62.1%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Lorenz Curve

45 degree line

Key Empirical Facts on Income Inequality

1) In the US, labor income inequality has increased substan-tially since 1970: debate between skilled biased technologi-cal progress view vs. institution view (min wage and Unions)[Autor-Katz’99]

2) In the US, top income shares dropped dramatically from1929 to 1950 and increased dramatically from 1980 to 2007[Piketty and Saez]

3) Top incomes used to be primarily capital income. Now, topincomes are divided 50/50 between labor and capital income(due to explosion of top labor incomes with stock-options,bonuses, etc.)

4) Fall in top income shares from 1900-1950 happened in mostOECD countries. Surge in top income shares has happenedprimarily in English speaking countries, not as much in Conti-nental Europe and Japan [Atkinson, Piketty, Saez JEL’11]

8

1940 1950 1960 1970 1980 1990 20000.30

0.35

0.40

0.45

0.50

Year

Gin

i coe

ffici

ent

● All WorkersMenWomen

●

●

● ● ●●

●

●● ●

●

●●

●●

●●

●● ●

●●

● ●● ● ● ● ●

● ● ●● ● ●

● ●●

● ●● ●

●

● ●

●●

● ●●

●

● ● ● ●

● ●●

●●

● ●●

●●

●●

●

●

●

● ● ●●

●

●● ●

●

●●

●●

●●

●● ●

●●

● ●● ● ● ● ●

● ● ●● ● ●

● ●●

● ●● ●

●

● ●

●●

● ●●

●

● ● ● ●

● ●●

●●

● ●●

●●

●●

●

Figure 1: Gini coefficient

Source: Kopczuk, Saez, Song QJE'10

25%

30%

35%

40%

45%

50%

1917

19

22

1927

19

32

1937

19

42

1947

19

52

1957

19

62

1967

19

72

1977

19

82

1987

19

92

1997

20

02

2007

Top

10%

Inco

me

Shar

e Top 10% Pre-‐tax Income Share in the US, 1917-‐2011

Source: Piketty and Saez, 2003 updated to 2011. Series based on pre-tax cash market income including realized capital gains and excluding government transfers.

0%

5%

10%

15%

20%

25% 19

13

1918

19

23

1928

19

33

1938

19

43

1948

19

53

1958

19

63

1968

19

73

1978

19

83

1988

19

93

1998

20

03

2008

Shar

e of

tota

l inc

ome

for e

ach

grou

p

Decomposing Top 10% into 3 Groups, 1913-2011

Top 1% (incomes above $367,000 in 2011)

Top 5-1% (incomes between $155,000 and $367,000)

Top 10-5% (incomes between $111,000 and $155,000)

Source: Piketty and Saez, 2003 updated to 2011. Series based on pre-tax cash market income including realized capital gains and excluding government transfers.

0%

2%

4%

6%

8%

10%

12% 19

13

1918

19

23

1928

19

33

1938

19

43

1948

19

53

1958

19

63

1968

19

73

1978

19

83

1988

19

93

1998

20

03

2008

Top

0.1%

Inco

me

Shar

e Top 0.1% US Pre-Tax Income Share, 1913-2011

Top 0.1% income share (incomes above $1.56m in 2011)

Source: Piketty and Saez, 2003 updated to 2011. Series based on pre-tax cash market income including or excluding realized capital gains, and always excluding government transfers.

Average Income Real Growth

Top 1% Incomes Real Growth

Bottom 99% Incomes Real

Growth

Fraction of total growth (or loss)

captured by top 1%

(1) (2) (3) (4)

Full period 1993-2011 13.1% 57.5% 5.8% 62%

Clinton Expansion 1993-2000 31.5% 98.7% 20.3% 45%

2001 Recession 2000-2002 -11.7% -30.8% -6.5% 57%

Bush Expansion 2002-2007 16.1% 61.8% 6.8% 65%

Great Recession 2007-2009 -17.4% -36.3% -11.6% 49%

Recovery 2009-2011 1.7% 11.2% -0.4% 121%

Computations based on family market income including realized capital gains (before individual taxes).Incomes exclude government transfers (such as unemployment insurance and social security) and non-taxable fringe benefits.Incomes are deflated using the Consumer Price Index.

Table 1. Real Income Growth by Groups

0

5

10

15

20

1910

19

15

1920

19

25

1930

19

35

1940

19

45

1950

19

55

1960

19

65

1970

19

75

1980

19

85

1990

19

95

2000

20

05

2010

Top

1% In

com

e Sh

are

(in %

) Top 1% share: English Speaking countries (U-shaped)

United States

United Kingdom

Canada

0

5

10

15

20

1910

19

15

1920

19

25

1930

19

35

1940

19

45

1950

19

55

1960

19

65

1970

19

75

1980

19

85

1990

19

95

2000

20

05

2010

Top

1% In

com

e Sh

are

(in %

) Top 1% share: Continenal Europe and Japan (L-shaped)

France

Japan

Sweden

Public Finance and Public Policy Jonathan Gruber Third Edition Copyright © 2010 Worth Publishers 6 of 34

C H A P T E R 1 7 ■ I N C O M E D I S T R I B U T I O N A N D W E L F A R E P R O G R A M S

17.1 Facts on Income Distribution in the United States Relative Income Inequality

According to these data, the share of income received by the lowest quintile in the United States is smaller than in any other nation, and is less than half of the worldwide average. The share of income received by the highest quintile in the United States is higher than in any nation except Mexico, and is nearly 25% higher than the worldwide average.

POVERTY RATE DEFINITIONS

1) Absolute: Fraction of population with disposable income

(normalized by family size) below poverty threshold z∗ fixed

in real terms (US case, World Bank $1/day in 1990 PPP)

2) Relative: Fraction of population with disposable income

(normalized by family size) below poverty threshold z∗ fixed

relative to median (European Union defines poverty threshold

as 60% of median)

Absolute poverty falls in the long run with economic growth [nobody inthe US is World Bank poor] but relative poverty does not

Absolute poverty captures both growth and inequality effects while relativepoverty captures only inequality effects

A recent study by Luttmer (2004) also finds that individuals’ self-reportedwell-being rises as their own income rises, but falls as their neighbors’incomes rise, suggesting that it is relative income, and not absolute income,that determines well-being.

14

Poverty Rate Disposable Income Definition

Most intuitive notion of poverty is based on consumption c

[not pre-tax income z]

c = z − T (z) + B(z) + E − s

where T (z) is tax, B(z) govt transfers, E net private transfers

(charity, family, friends), s is net savings (change in assets)

Consumption c is difficult to measure

Disposable Income z − T (z) + B(z) [post-tax income] mea-

sured in traditional Current Population Survey (CPS) [but does

not fully capture in-kind elements of B(z) such as Medicaid]

15

FAMILY SCALE

Ideally, poverty should be defined at the individual level based

on individual consumption [e.g., kids better off when mother

or grandmother controls income instead of father]

However, many consumption goods are public goods within

the family [e.g., housing, joint meals, etc.] and hard to mea-

sure consumption at individual level

Measured poverty is therefore based on consumption or dis-

posable income at the family level [or unit sharing resources]

and everybody within the family has same poverty status

Bigger families need more resources but economies of scale in

consumption: scale disposable income by family size

16

US POVERTY RATE DEFINITION

http://www.census.gov/hhes/www/poverty/poverty.html

Definition developed in 1963 by Molly Orshansky (at Social

Security Administration) as 3 times amount required to buy a

“thrifty food plan”

[3 times because 1955 survey showed that low income families

spent 1/3 of income on food]

Poverty threshold afterward indexed on official Consumer Price

Index and adopted by Administration as Official Poverty Index

Family scaling was based on 1955 survey

17

US POVERTY RATE DEFINITION

Based on money income = market income before taxes +

cash govt transfers + cash private transfers

In-kind market income and transfers (employer health insur-

ance, Medicaid, nutrition, public housing) do NOT count

Income and employee payroll taxes are NOT deducted, Income

tax credits (EITC, Child Tax Credit) are NOT added

Threshold depends on household size/structure: e.g., $18K/year

for single parent with 2 kids

Thresholds adjusted annually using the official CPI

18


C H A P T E R 1 7 ■ I N C O M E D I S T R I B U T I O N A N D W E L F A R E P R O G R A M S

17.1 Facts on Income Distribution in the United States Absolute Deprivation and Poverty Rates

ISSUES WITH US POVERTY RATE DEFINITION

Definition was close to disposable income in 1963 but no

longer:

1) In-kind transfers have grown substantially [Medicaid]

2) Payroll tax and Income tax credits (EITC, Child Tax Credit)

have grown substantially at the bottom

3) CPI is for the average consumer, low income consumer price

might evolve differently (Walmart) [also geographical varia-

tion]

Politically difficult to change definition

20

declined steadily during this period, falling from 24.6 percent in 1970 to10.2 percent in 2003.

Other factors may better explain why the poverty rate has failed to fall. Risingnumbers of female headed families may offset income gains from women’s increas-ing labor force participation. Increasing income inequality—in particular stem-ming from declines in wages for less-skilled workers—may have limited the poverty-fighting effects of economic growth. Finally, the level of and changes ingovernment benefits directed toward the nonelderly may explain why the noneld-erly poverty rate has not moved in the same direction as elderly poverty. Our taskin this paper is to document and quantify the effects of these competing factors tounderstand recent poverty trends better. Since the steady fall in elderly povertyrates in recent decades is likely explained by other factors such as Social Security(Englehardt and Gruber, 2004), we focus throughout this paper on the conundrumof why the nonelderly poverty rate has failed to decline as the economy hasexpanded.

Dimensions of Poverty

In this section, we summarize some basic facts about poverty in the UnitedStates, relying on a combination of previously published data from the Census

Figure 1Trends in Individual Poverty Rates and Real GDP per Capita, 1959–2003

40

35

30

25

20

15

10

5

0

40,000

35,000

30,000

25,000

20,000

15,000

10,000

5,000

0 1963

AllNonelderlyChildren

1968

Pove

rty

rate

GD

P pe

r ca

pita

(20

03$)

1973 1978 1983 1988 1993 1998 2003

ElderlyGDP per capita

Source: Poverty rates are from U.S. Bureau of the Census, Current Population Survey, Annual Socialand Economic Supplements. The GDP per capita series is from the Economic Report of thePresident (2005).Note: The poverty rate data are unavailable for some subgroups for 1960–1965.

48 Journal of Economic Perspectives

FACTORS EXPLAINING EVOLUTION OFPOVERTY

Based on Hoynes-Page-Stevens JEP’06

1) Increasing pre-tax inequality: stagnant bottom wages inspite of economic growth per capita [large effect]

2) Changes in family structure: single parent families ↑ from7% in 1967 to 14.4% in 2003 ⇒ Increases poverty rate by 4pts [large effect]

3) Increase in female labor force participation ⇒ Reducespoverty rate [significant effect only since 1980]

4) Immigration: accounts for about 0.7 points in the povertyrate increase from ’69 to ’99 [small effect]

5) Means-tested transfers [medium effect because they areconcentrated below poverty line]

22

parent families comprise 39.1 percent of the poor, although persons in suchfamilies make up only 14.4 percent of the total population.

The racial and ethnic composition of the poor is disproportionately minority,but the modal poor individual is a white non-Hispanic. In 2003, 42.2 percent of thepoor were white, 24.1 percent black and 26.8 percent Hispanic. In the overall popu-lation, whites make up 65.7 percent, blacks make up 12.6 percent, and Hispanics15.1 percent. Immigrants are 17.4 percent of the poor. The bottom row of Table 1shows that half of the poor were in a family whose household head worked in the pastyear. In the population overall, 81 percent of household heads worked.

Persistence of PovertyOne dimension of poverty that cannot be captured using data from the

Current Population Survey is its persistence, since the CPS only asks about incomein a given year and does not ask about individuals’ income history. Bane andEllwood (1986) provide a fundamental contribution to our understanding of thedynamics of poverty. In particular, imagine that during a calendar year one familyis poor for all 12 months and 12 other families are poor for only one month each.At any given time, two families are poor, and half of those who are poor at any giventime are poor for the long term. But over the course of a year, only one of the13 families who experienced at least one month of poverty were poor for an

Table 1Characteristics of the Nonelderly Poor, 2003(percentage with given characteristic)

Among nonelderly poor Among all nonelderly

Individual characteristicsAge �18 39.8% 28.8%Male 45.5% 49.8%Female 54.5% 50.2%Family head is

Married 35.0% 66.6%Single with kids 39.1% 14.4%Single without kids 25.8% 18.9%

White 42.2% 65.7%Black 24.1% 12.6%Hispanic 26.8% 15.1%Family head’s education

�High school 35.3% 14.4%Native-born 82.6% 87.4%Immigrant 17.4% 12.6%Head worked last year 50.0% 81.1%

Source: Author’s tabulations of the 2004 March CPS.Note: The age, gender, race and ethnicity are assigned using the individual’s characteristics. Family type,immigrant status, education and employment are assigned based on characteristics of the head of the family.


Figure 2Nonelderly Poverty Rates, Unemployment Rates and Median Wages, 1967–2003

0.20 Poverty rateUnemployment rateReal median wage

0.15

0.10

0.05

Un

empl

oym

ent r

ate,

pov

erty

rat

e

Med

ian

rea

l wee

kly

earn

ings

(20

03$)

0.00 500

550

600

650

700

750

800

850

900

950

1000

1967 1971 1975 1979 1983 1987 1991 1995 1999 2003

Source: Authors’ tabulations of the 1968–2004 March CPS.Notes: Median hourly wages are defined for all full-time working men. See text for more details.

Figure 3Nonelderly Poverty Rates and Inequality, 1967–2003

0.200 Nonelderly poverty rateMedian wage/20th percentile wage

0.175

0.150

0.125

0.100

Pove

rty

rate

Med

ian

wee

kly

wag

e/20

th p

erce

nti

le w

eekl

y w

age

0.075

0.050

2.000

1.750

1.500

1.250 1967 1971 1975 1979 1983 1987 1991 1995 1999 2003

Source: Authors’ tabulations of the 1968–2004 CPS.Notes: Inequality is measured as the ratio of median weekly wage to the 20th percentile weekly wage.Wages are defined using all full-time working men. See text for more details.


graphic changes can explain trends in the poverty rate (Cancian and Reed, 2001;Blank and Card, 1993). Here we update that literature.

Table 3 presents the results of our analysis. The first two columns of Table 3show the distribution of individuals in 1967 and 2003, by family type. We categorizeindividuals by one of six different family types: married individuals with and withoutchildren; single females with and without children; and single males with andwithout children. Table 3 shows that in 2003, 67 percent of persons lived in marriedcouple families, down from 86 percent in 1967. In contrast, the percentage ofpersons living in unmarried parent families increased from 7 percent in 1967 to14.4 percent in 2003. In columns 3 and 4, we provide the actual poverty rates forpersons in each family type. While poverty rates decreased between 1967 and 2003for all groups, there are persistent differences across groups—with the highestpoverty rates for persons in single parent families and the lowest poverty rates forpersons in married couple families.

We can use these data to illustrate the change in poverty between 1967 and2003 that is predicted purely from changes over time in the fraction of individualsliving in different family types. Specifically, we hold constant the poverty rateswithin each family type at their 1967 level, but allow the fraction of individualsliving in each family type to change to their 2003 levels. Changes in family structurealone predict that poverty rates should have risen from 13.3 percent in 1967 to17 percent in 2003. Thus, like the changes in unemployment, median wages andwage inequality, changes in family types substantially overpredict the actual in-crease in poverty rates over time.

How were the higher poverty rates predicted by the population shift toward

Table 3Effect of Family Structure on Nonelderly Poverty Rates

Percentage ofnonelderly persons by

family type

Percentage ofnonelderly persons inpoverty by family type

1967 2003 1967 2003

Persons by family typeMarried couples with children 67.3 44.2 10.7 8.1Married couples without children 18.7 22.4 5.8 4.1Single women with children 6.2 11.9 51.2 37.3Single men with children 0.8 2.5 28.4 22.0Single women without children 4.4 9.6 25.4 18.6Single men without children 2.6 9.3 18.1 16.2

All personsPercentage in poverty, actual 13.3 12.8Predicted poverty, changes in family type only 17.0

Source: Authors’ tabulations of the 1968 and 2004 March CPS.


have a larger effect than means-tested cash payments, reducing the poverty rate bynearly 3 percentage points from 15.2 to 12.4 percent.

The Bureau of the Census also provides calculations of income and povertythat include noncash transfers, which are based on assumptions about the cashequivalent value of each in-kind benefit program. The impacts on poverty areshown in lines (h), (i) and (j) of Table 4. Comparing lines (h) and (j), we see thatmeans-tested noncash transfers reduce poverty by about 1.5 percentage points.

Taken together, these calculations suggest that government programs do havea modest effect on poverty, even though many of them are not accounted for in theofficial rate. More to the point, these programs may have a substantial effect on the

Table 4Percentage of Persons in Poverty by Alternative Definition of Income, 2003,Measuring Impacts of Government Programs

Nonelderlypersons Children

(a) Official poverty measure(Money income � pretax, postgovernment cash transfers) 12.7 17.6Poverty reduction due to EITC(b) Money income (official measure) less all taxes except EITC 13.9 19.1(c) Money income less all taxes (including EITC) 12.2 16.0Poverty reduction due to means-tested cash transfers(d) Full income less taxes less means tested government cash transfersa 12.2 15.8(e) Full income less taxes 11.4 14.9Poverty reduction due to non means-tested cash transfers(f) Pregovernment transfer money income less taxesb 15.2 17.8(g) Pregovernment transfer money income less taxes plus nonmeans

tested cash government transfers12.4 15.9

Poverty reduction due to means-tested noncash transfers(h) Full income less taxes (definition e above) 11.4 14.9(i) Full income less taxes plus Medicaid 10.8 13.8(j) Full income less taxes plus Medicaid plus other means-tested

government noncash transfers9.9 12.3

Source: U.S. Bureau of the Census (2005) and special tabulations by the Census Bureau.Notes: To locate these figures in the Census report, note that (a) is Census definition 1; (b) is Censusdefinition 1a; (c) is Census definition 1b; (d) is Census definition 11; (e) is Census definition 12; (f) isCensus definition 8; (g) is Census definition 9; (i) is Census definition 13; and (j) is Census definition14. Taxes include payroll taxes, federal and state taxes. Means-tested government cash transfers includeTANF, Supplemental Security Income, means tested Veteran’s payments and other public assistance.Non-means-tested government cash transfers includes Social Security, unemployment compensation,worker’s compensation, nonmeans tested Veteran’s payments, Railroad Retirement, Black Lung pay-ments, Pell Grants and other educational assistance. Means-tested noncash transfers include foodstamps, rent subsidies, and free and reduced-price school lunches. For details on simulating taxes, seeO’Hara (2004). For details on calculating the value of noncash benefits, see U.S. Bureau of the Census(1992).a Full income includes pretransfer money income less means tested transfers plus capital gains, em-ployer paid health insurance, Medicare and regular-price school lunches.b Income measure also includes capital gains and employer paid health insurance.

Poverty in America: Trends and Explanations 63

the poverty rate fell so much between 1959 and 1969, while a growing andincreasingly low-income immigrant population cannot explain much of the trendin poverty prior to 1980. On the other hand, if we focus on the second half of theperiod, we see that while poverty rates among natives have changed little, povertyrates among immigrants have increased by nearly two percentage points, and thefraction of the population that is foreign born has increased by six percentagepoints. Taken together, these changes should put upward pressure on the povertyrate, but how much?

To answer this question, we begin by considering the extent to which overallpoverty would have declined if the share of immigrants had increased over time butimmigrants and natives had kept same poverty rates as in 1979. We find that if thelevel of poverty among immigrants had stayed the same as it was in 1979, the risingshare of immigrants would have increased the poverty rate from 12.3 percent(1979) to 12.5 percent (1999), a number that is only slightly bigger than the actualvalue of 12.4 percent. We also consider the effects of changes over time in thefraction of immigrants who are poor. If we hold population shares and nativepoverty rates constant at their 1979 levels, but allow poverty rates among immi-grants to vary across Census years, then the predicted overall poverty rate in 1999is about 0.1 percentage points higher than its 1979 level. Although recent immi-grants are poorer than their predecessors, their fraction of the population is simplytoo small to affect the overall poverty rate by much.

These calculations are based on an important assumption, however, which isthat large influxes of immigrants do not reduce job opportunities available tonatives. If the presence of immigrant workers depresses native’s wages, then theoverall impact of immigration on the poverty rate will be higher. Evidence on thelabor market effects of immigration is mixed (see Borjas, 1999, for an overview ofthis literature), but it seems safest to consider these estimates as lower bounds.

Table 5Nonelderly Poverty Rates in Native and Immigrant Households, by Year

All personsPersons in households headed

by a nativePersons in households headed

by an immigrant

Poverty rate Poverty ratePercentage ofpopulation Poverty rate

Percentage ofpopulation

1959 20.6 20.9 95.8 14.1 4.21969 12.4 12.5 95.9 11.2 4.11979 12.3 12.1 94.0 15.6 6.01989 12.9 12.5 91.4 17.5 8.61999 12.4 11.8 87.9 17.4 12.1

Source: Authors’ tabulations of 1960, 1970, 1980, 1990 and 2000 Census files.

Hilary W. Hoynes, Marianne E. Page and Ann Huff Stevens 65

Recomputing Poverty Rate: Meyer-Sullivan NBER’09

1) Change the scaling for family size (no strong effect)

2) Change the price index: shift to CPI-U-RS instead of officialCPI-U (large effect, controversial, a problem with absolutepoverty thresholds)

3) Shift to households [people living in same unit] instead offamily [people in same unit related by blood/adoption]: notclear which is best, depends on sharing [some effect]

4) Shift to after-tax income [deduct income/payroll taxes, addtax credits]: large legitimate effect

5) Add non-cash benefits [nutrition, housing, health insur-ance]: tiny net effect [medicaid ↑, other programs ↓]

6) Shift to consumption [modest effect on poverty rate, hugeeffect on deep poverty]

24

Notes: The rates are anchored at the official rate in 1980. Data are from the CPS-ASEC/ADF. Official Income Poverty follows the U.S. Census definition of income poverty using official thresholds. For measures other than the official one, the threshold in 1980 is equal to the value that yields a poverty rate equal to the official poverty rate in 1980 (13.0 percent). The thresholds in 1980 are then adjusted overtime using the CPI-U-RS. Poverty status is determined at the family level and then person weighted. After-Tax Money Income includes taxes and credits (calculated using TAXSIM). After-Tax Money Income + Noncash Benefits Excluding Home Equity also includes food stamps and CPS-imputed measures of housing and school lunch subsidies, and the fungible value of Medicaid and Medicare. This last series is only available starting with the 1980 CPS-ASEC/ADF. See Data Appendix for more details.

0.07

0.08

0.09

0.1

0.11

0.12

0.13

0.14

0.15

0.16

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

Frac

tion

Poor

Official Income Poverty (CPI-U)

Money Income (NAS Scale, CPI-U-RS)

After-Tax Money Income (NAS Scale, CPI-U-RS)

After-Tax Income + Noncash Benefits Excluding Home Equity (NAS Scale, CPI-U-RS)

Figure 1: Official and Alternative Income Poverty Rates, 1972-2005

Rates anchored at the official rate in 1980

Notes: The rates are anchored at the official rate in 1980. Poverty status is determined at the family level and then person weighted. Consumption data are from the CE Survey and income data are from the CPS-ASEC/ADF. Official Income Poverty and After-Tax Money Income Poverty are as in Figure 1. CE Survey data are not available for the years 1974-1979 and 1982-1983. Also, consumption data are not available for the years 1984-1987 for measures that include health insurance.

0.07

0.08

0.09

0.1

0.11

0.12

0.13

0.14

0.15

0.16

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

Frac

tion

Poor

Official Income Poverty (CPI-U)

After-Tax Money Income (NAS Scale, CPI-U-RS)

Consumption (NAS Scale, CPI-U-RS)

Consumption Excluding Health Insurance (NAS Scale, CPI-U-RS)

Figure 2: Consumption and Income Poverty Rates, 1972-2005

Govt Redistribution with Taxes and Transfers

Govt taxes individuals based on income and consumption and

provides transfers: z is pre-tax income, y = z − T (z) + B(z) is

post-tax income

1) If inequality in y is less than inequality in z then tax and

transfer system is redistributive (or progressive)

2) If inequality in y is more than inequality in z then tax and

transfer system is regressive

a) If y = z · (1− t) with constant t, tax/transfer system is neutral

b) If y = z · (1− t) + G where G is a universal (lumpsum) allowance, thentax/transfer system is progressive. Actual tax/transfer systems in OECDcountries look roughly like this.

c) If y = z−T where T is a uniform tax (poll tax), then tax/transfer systemis regressive

26

Federal US Tax System: Overview

1) Individual income tax (on both labor+capital income) [pro-

gressive](40% of fed tax revenue)

2) Payroll taxes (on labor income) financing social security

programs [about neutral] (40% of revenue)

3) Corporate income tax (on capital income) [progressive if

incidence on capital income] (15% of revenue)

4) Estate taxes (on capital income) [very progressive] (2% of

revenue)

5) Minor excise taxes (mostly labor income) [regressive] (3%

of revenue)

27

State+Local Tax System: Overview

1) Individual+Corporate income taxes [progressive] (30% of

state+local tax revenue)

2) Sales + Excise taxes (tax on consumption = income -

savings) [about neutral] (30% of revenue)

3) Real estate property taxes (on capital income) [slightly pro-

gressive] (30% of revenue)

http://www.census.gov/govs/www/qtax.html

28


C H A P T E R 1 8 ■ T A X A T I O N I N T H E U N I T E D S T A T E S A N D A R O U N D T H E W O R L D

18.1 Types of Taxation Taxation around the World

US Tax System: Progressivity and Evolution

1) Medium Term Changes: Federal Tax Progressivity has

declined since 1970 but govt redistribution remains substantial

especially when including transfers (Medicaid, Social Security,

UI, DI, various income support programs)

2) Long Term Changes: Before 1913, US taxes were pri-

marily tariffs, excises, and real estate property taxes [slightly

regressive], no transfer programs (and hence small govt)

http://www.treasury.gov/education/fact-sheets/taxes/ustax.shtml

30

2. Federal Average Tax Rates by Income Groups (individual+corporate+payroll+estate taxes)

0%

10%

20%

30%

40%

50%

60%

70%

80%P0

-20

P20-

40

P40-

60

P60-

80

P80-

90

P90-

95

P95-

99

P99-

99.5

P99.

5-99

.9

P99.

9-99

.99

P99.

99-1

00

1970

2000

2005

Source: Piketty and Saez JEP'07

Plan for Lectures on Taxation/Redistribution

1) Tax incidence (who bears the burden of taxation), efficiency

costs of taxation, optimal commodity taxation

2) Taxation of labor income:

Optimal design of labor income taxation and means-tested

transfers

Empirical analysis of tax and transfer programs on labor supply

and earnings

3) Taxation of capital income (savings, wealth, and corporate

profits)

32

introduction to taxes and transfers: income distribution ...saez/course131/taxintro_ch17.pdf ·...

Documents