The Distribution Function of Government:

(Chapters 12 and 13)

Recall… Government Intervention in markets and the

economy is most often justified by arguing for government
to serve one of the three following functions:
1. Allocation Function – (discussed in previous topic)
2. Distribution Function – government policies aimed at
changing the final distribution of goods/services across
consumers, usually with the intention of realizing a
“fairer” apportionment of consumption/income/wealth.
3. Stabilization Function – (discussed in next topic)

“Redistribution” – government policies designed to alter

the distribution of income/wealth/consumption across
members of society
 Income – the flow of money earned by an individual
during a period of time
 Wealth – the current stock of money (and other valuable
assets) that an individual owns at a point in time

Broadly two different ways to directly “redistribute”

1. Income Support – monetary payments (such as
unemployment benefits and social security payments)
made to certain individuals, which directly alter the
distribution of income within a society.
2. Redistribution In-Kind – public provision of
goods/services (such as healthcare, education,
housing, food) for certain individuals, which alter the
consumption of goods/services within society.
131
Redistribution policies often intended to reduce poverty…
 Poverty – the condition of having very limited access to
goods and services
 Poverty thresholds – income levels (based upon the
number of and ages of individuals in a household) below
which a household is deemed to be living in poverty
 Poverty rate – the percentage of the population that is
living in poverty

In many respects, poverty is a relative concept => what it

means to be living in poverty in the U.S. is drastically
different than what it means to be living in poverty in Togo
 World Bank sets their global poverty threshold at an
income of $1.90 per day (i.e., $693.50 per year). Global
Poverty rates…
 1981: 44.3% (2 billion out of 4.5 billion)
 2015: 9.6% (700 million out of 7.3 billion)
 By the standard of $1.90 per day, virtually nobody in the
U.S. is living in poverty

 Poverty thresholds in U.S. (2015)

 $12,331 for a single person under age 65
 $24,036 for a family of four with two adults and two
children

132
Poverty Rate in the U.S., 1959-2014

 Maximum rate over this period: 22.4% in 1959

 Minimum rate over this period: 11.1% in 1973
 Equal to 14.8% in both 2013 and 2014

Measuring poverty is complicated by the facts that:

(i) we continually “raise the bar” for what it means to
be not living in poverty
(ii) Due to improvements in technology and general
economic growth, overall standards of living have
greatly increased in recent decades
All Households in 1950 versus Poor Households in 1997
Household Ownership Rate for All Ownership Rate for Households
Amenity Households in 1950 Below Poverty Line in 1997
Electricity 94% 99%
Flush Toilet 76% 99%
Refrigerator 80% 99%
Television 10% 95%
Telephone 79% (1960) 88%
Automobile 59% 71%
Air Conditioning 12% (1960) 70%
Washing Machine 47% 58%
Clothes Dryer 2% 48%
Dishwasher 2% 28%

133
Thinking about redistribution, such policies are based on
the premise that the distribution of income, wealth, and
consumption resulting from the market system is not best
 But, “what distribution of income is best or most
desirable?” Is a “more even distribution of income”
necessarily an “ideal” for which we should strive?
 What determines levels of income/wealth (and
therefore consumption) in a free market system?

7 Determinants of Productivity, Income, and Wealth:

1. Natural Talent and Ability – ability is not distributed
equally at birth; some people possess attributes (e.g.,
intelligence, strength) that make them more productive
2. Acquired Skills – individual productivity depends in
large part upon skills and experiences acquired during
education, training, and work experience
3. Effort – productivity is often largely dependent upon
effort (“workers” are more productive than “shirkers”)
4. Compensating Differentials – differences in wage
rates that are due to differences in working conditions
(e.g., risky jobs pay more; “glamour jobs” pay less)
5. Inherited Wealth – wealth is distributed less evenly
than income (due partly to differences in inheritances)
6. Accumulated Savings – the stock of wealth that a
person has at any point in time is partly determined by
previous consumption/savings decisions
7. Seemingly Unrelated Market Conditions – income
depends in part upon economic conditions beyond our
control; a worker’s value depends greatly upon the
price of the product he helps produce (e.g., wage of a
coal miner would increase if we ran out of oil)
134
 How is income distributed in the U.S.?
Income Inequality in the U.S. (2013), based on Income Tax data
Bottom Bottom Bottom Bottom Bottom
Segment of the Population
50% 75% 90% 95% 99%
Cumulative Share of Income 11.5% 31.9% 54.1% 65.6% 81.0%
Income Cutoff $36,841 $74,955 $127,695 $179,76 $428,713
0

Can we graphically illustrate the distribution of income

across members of society? Can we develop a single
numerical measure of income inequality?

 Lorenz Curve – a graph (developed by economist Max

Lorenz) which illustrates income inequality within a
society by plotting the relation between cumulative
fraction of the population (with people ordered from
lowest income to highest) and corresponding cumulative
fraction of total income earned.
 “More even” distribution of income  Lorenz Curve
“closer to 45-degree line”
 Gini-Coefficient – a quantitative measure of income
inequality based upon the Lorenz Curve, defined as the
ratio of “the area between 45-degree line and the Lorenz
Curve” to “the entire area below 45-degree line.”

135
 Lorenz Curve and Gini-Coefficient:
0
Fraction of “ 45 – Line” or
Total Income “Line of Perfect
Income Equality”
1

.810

.541

.319 Lorenz Curve

for U.S. (2007)
.115

0 Fraction of Total
Population
0
.50 .75 .90 .99 1

Green
(Gini-Coefficient) =
( Green+Blue )
Fraction of Total Income

0
“ 45 – Line” or
“Line of Perfect
Income Equality”

Lorenz Curve
for U.S. (2007)

0 Fraction of Total
1 Population
0
136
  Extreme Equality => Lorenz Curve on 45-degree line
0
G= 1
=0
=> Gini-Coefficient equal to zero: 0+ 2
Fraction of Total Income
1

0 Fraction of Total Population

0 1

 Extreme Inequality (one person having all income) =>

Lorenz Curve lies along horizontal axis => Gini-
1
2
G= 1 =1
Coefficient equal to one: 2
+0
Fraction of Total Income
1

0 Fraction of Total Population

0 1

Year 1974 1984 1994 2004 2014

U.S. Gini Coefficient .395 .415 .456 .466 .480

 The distribution of incomes in the U.S. has become

“less equal” over the past four decades

137
 Question: Even if you “value” income equality, should you
“always prefer” a smaller value of the Gini-Coefficient?
 Recognize that the Gini-Coefficient tells us nothing
about “income levels,” but rather only describes “how an
amount of income is divided”
 Average Real (adjusted for inflation) Household Income
generally increased over most of this period
Year 1974 1984 1994 2004 2014
U.S. Gini Coefficient .395 .415 .456 .466 .480
Median Household Income $48,497 $48,664 $51,006 $55,565 $53,657
Mean Household Income $56,713 $59,625 $68,189 $75,784 $75,738
 Looking at changes in income levels for different
segments of society…
Mean Household Income for Different Segments of the U.S.
Year 1974 1984 1994 2004 2014
Bottom 5th $12,053 $11,802 $12,195 $12,839 $11,676
Second Lowest 5th $29,957 $29,395 $30,391 $32,852 $31,087
Middle 5th $48,282 $48,713 $51,198 $55,661 $54,041
Second Highest 5th $69,728 $73,347 $79,670 $87,766 $87,834
Top 5th $123,54 $134,86 $167,48 $189,80 $194,053
2 6 9 2

 If you were going to be “randomly placed” somewhere in

the income distribution in either 1974, 1984, 1994, or
2004 (i.e., ignore 2014), which one would you pick?
 To me, the obvious best choice is 2004, even though it
has the highest valued Gini Coefficient (2004 income

138
 levels in all five quintiles “dominate” corresponding
levels in 1974, 1984, and 1994)

4 Arguments in Favor of Coercive Redistribution:

1. Utilitarian Justice – total social welfare can be

increased by transferring income/wealth from the rich
to the poor, so long as people have a diminishing
marginal utility for money
 idea that society should strive for the “greatest
happiness for the greatest number of people” (i.e.,
attempt to maximize the summation of “utility” or
“happiness” over all members of society)
 arguments first put forth by Jeremy Bentham (1748-
1835) and John Stuart Mill (1806-1873)
 if the value of $1 is greater if you have less money
(i.e., diminishing marginal utility), then social
welfare can be increased by taking $1 away from
Ted Turner and giving it to a homeless person

139
 Jeremy Bentham's Auto-Icon at University College London
2. Rawlsian Justice (Social Contract Theory) – the
socially best income distribution is the one which
maximizes the well-being of the worst-off member of
society
 developed by the philosopher John Rawls in A
Theory of Justice (1971)
 consider the “ideal society” that would be designed
by someone in the “beforelife” behind a “veil of
ignorance” (i.e., not knowing what their realized lot
in life would be, but just knowing that it would be a
random draw in this “ideal society”)
 Rawls argues that such a person would be
particularly concerned about the situation of the
poorest person (i.e., the worst possible outcome)

140
 Therefore, he argues that this objectively ideal
society should implement policies to apply the
“maximin criterion”
 Maximin Criterion – a claim that the government
should aim to maximize the well-being of the worst
off person in society

141
3. Labor Theory of Value – an assessment of the
production process which attributes all economic
surplus generated from production to labor
 based upon the ideas of Karl Marx (1818-1883)
 mainstream economists do not agree with this
outdated notion => rather, the value of an item
clearly depends upon benefits to consumers
 e.g., if Ann spends 4 hours per day growing corn
and Bob spends 8 hours per day digging and then
filling ditches, is Bob’s output twice as valuable
 recognize that Marx’s notion attributes no value to
the capital (and other “non-labor inputs”) necessary
for the production process
 since a capitalist system provides returns for all
inputs (i.e., not only labor, but also capital and
land), labor is vastly underpaid relative to what
Marx believed was “fair”
 thus, in the interest of fairness, income/wealth must
be redistributed away from the owners of capital
and toward workers

4. Overcoming the Free-Rider Problem

 suppose society prefers a more equal distribution of
income/wealth than what the market gives
 if we rely upon private charity to reduce inequality,
there is a free rider problem (recall the discussion
from market provision of public goods)
 thus, private charity does not provide enough
redistribution => have government redistribute, to
get an amount closer to the socially best level

142
An Argument Against Coercive Redistribution:

5. Libertarian Justice – the argument that the fairest

distribution of income and consumption is that which
results when the government establishes and enforces
a legal code which respects all voluntary economic
interactions between individuals in society
 many of these ideas best articulated by Robert
Nozick in Anarchy, State, and Utopia (1974)
 government sets rules of the “game” which place
the utmost respect on individual property rights and
choices, and then simply acts as the “referee”
 “society” does not earn income; only individuals
within society earn income => society has no
rightful claim to the income earned by any person
 whereas others try to judge “what distribution of
income is best?,” Nozick and the proponents of
Libertarian Justice deny the validity of the question
 as long as the process for determining incomes is
fair and just, the resulting distribution is fair and
just (no matter how equal or unequal)
 equality of opportunity is more important than
equality of outcomes
 government should enforce laws to guarantee that
all people have the same opportunities to use
their talents to earn income => but once the rules
are established, the government has no reason (in
fact, no legitimate authority) to alter the resulting
distribution of income

143
 Redistribution Programs and Policies in the U.S.:
2012: U.S. government spending on welfare programs…
 roughly $952 billion => about 6% of GDP
 about $20,610 per poor person in the country
 so high partly because, due to the complex
bureaucracy in place, it takes about $7 of taxes to get
$1 of additional income into the hands of a recipient

Most Costly U.S. Federal Government Welfare Programs

Program Annual Cost Number of Participants
Medicaid $228.0 billion 48,900,000
Supplemental Nutrition Assistance Program $75.0 billion 44,200,000
Earned Income Tax Credit $55.0 billion 27,000,000 households
Supplemental Security Income $43.7 billion 8,100,000
Federal Pell Grants $41.0 billion 9,614,000
Temporary Assistance for Needy Families $21.0 billion 4,492,000 (monthly avg.)
Section 8 Housing Choice Vouchers $18.1 billion 2,000,000 households
Very Low to Moderate Income Housing Loans $16.7 billion 131,370 (units)
Title 1 Grants to Local Education Agencies $14.1 billion n.a.
Children’s Health Insurance Program $13.5 billion 7,705,723
National School Lunch Program $10.9 billion 31,000,000
Adjustable Rate Mortgage Program $10.6 billion 43,687 (units)
Maternal/Infant/Childhood Home Visiting Program $7.5 billion n.a.
Supplemental Nutrition for Women/Infants/Children $7.2 billion 9,180,000
Head Start $7.1 billion 904,000

 A mix of “income support” and “redistribution in-kind”

*** table above does not even include the two biggest
Federal government redistribution programs ***
 Social Security: $888 billion to roughly 40 million people
 Medicare: $546 billion to roughly 55 million people

144
Altering the Distribution of Income via Taxation:
 We can essentially redistribute income from the rich to
the poor by taxing the rich more heavily
 Common to describe a “tax structure” as Progressive,
Proportional, or Regressive, by examining how the
“Average Tax Rate” behaves as income is increased
 Average Tax Rate (ATR) –the amount of total taxes
paid divided by income
i. Progressive Tax – tax for which ATR increases as
the level of income is increased.
ii. Proportional Tax – tax for which ATR remains
constant as the level of income is increased.
iii. Regressive Tax – tax for which ATR decreases as
the level of income is increased.
 Marginal Tax Rate (MTR) –the percentage of the
next dollar earned that must be paid in taxes.
 As a general rule (i.e., mathematical implication):
1. ATR will increase whenever MTR> ATR
2. ATR will decrease whenever MTR< ATR
Examples:
i. Progressive: U.S. Federal Income Tax
ii. Proportional: “flat tax” with no deductions whatsoever
[e.g., “flat tax” proposed by Steve Forbes in 1996 was
NOT a proportional tax, since the first $33,000 of
earnings were not taxed; income taxes in Bulgaria
(10%), Hungary (16%), Iraq (15%), Jamaica (25%),
Romania (16%), Russia (13%), and Ukraine (15%)]
iii. Regressive: Social Security Payroll Tax – first
$118,500 taxed at a marginal rate of 6.2%, while
additional earnings are not taxed at all
145
  a “highly progressive tax” places a disproportional
amount of the tax burden on high income earners
 measuring income inequality by the Gini Coefficient, a
progressive tax reduces income inequality (i.e., imposing
the tax decreases the value of the Gini Coefficient)
 The U.S. Federal Income Tax is a progressive tax which
reduces income inequality
 Questions…
1. “How progressive” is the U.S. Federal Income Tax?
2. “How much” does the U.S. Federal Income Tax
redistribute income?

U.S. Federal Income Tax…

 16th Amendment to the U.S. Constitution ratified on
2/3/1913 – authorized congress to levy an income tax
 U.S. Federal Income Tax established in 1913
 Progressive Tax (increasing MTR)
 Primary source of U.S. Federal Government revenue:
$1.541 trillion of revenue in 2015 (47.43% of all Federal
revenue)
Tax Brackets and % of Population Filing Taxes over time
Year 1913 1916 1921 1931 1933 1941 1945
High MTR 7% 15% 73% 25% 63% 81% 94%
Low MTR 1% 2% 4% 1.5% 4% 10% 23%
Number of Brackets 7 14 56 23 55 32 24
% of Adult Pop. On Returns 1.1% 1.2% 15.0% 5.7% 6.6% 42.9% 75.9%
Year 1963 1964 1981 1990 2000 2010 2015
High MTR 91% 77% 70% 28% 39.6% 35% 39.6%
Low MTR 20% 16% 0% 15% 15% 10% 10%
Number of Brackets 24 26 16 2 5 6 7
% of Adult Pop. On Returns 83.6% 84.6% 83.9% 86.5% 84.9% 82.1% n.a.

146
 Current Marginal Tax Rates, U.S. Federal Income Tax
Range of Adjusted Gross Income Marginal Tax Rate
$0 up to $18,150 10%
$18,151 up to $73,800 15%
$73,801 up to $148,850 25%
$148,851 up to $226,850 28%
$226,851 up to $405,100 33%
$405,101 up to $457,600 35%
$457,601 and above 39.6%

Increasing MTRs => ATR rises with AGI => Progressive

 But, how exactly is tax burden distributed over different
income groups?

(i) Tax revenues disproportionately comes from ‘the rich’

% of Adjusted Gross Income & Income Taxes Paid
(2013)
Segment of Lower Income Group’s Share of Total Group’s Share of Total
Population Threshold Income Taxes Paid Adjusted Gross Income
Top 1% $428,713 37.80% 19.04%
Top 5% $179,760 58.55% 34.42%
Top 10% $127,695 69.80% 45.87%
Top 25% $74,955 86.27% 68.10%
Top 50% $36,841 97.22% 88.51%

 “Top 5%” pay 58.55% of all income tax dollars

 “Bottom 95%” pay (100–58.55) = 41.45%
 “Top 10%” pay 69.80% of all income tax dollars
 “Bottom 90%” pay (100–69.80) = 30.2%
 “Top 1%” pay more than “Bottom 90%”: 37.8% v 30.2%
 “Top 1%” pay almost as much as “Bottom 95%”: 37.8%
v 41.45%
 “Top 50%” pay 97.22% of all income taxes
147
 “Bottom 50%” only pay 2.78% of all income taxes

148
 (ii) Average Tax Burden has evolved in recent decades,
but U.S. Federal Income Tax clearly is (and has been)
progressive
Average Tax Rates Over Time
Between Between Between Bottom
Year Top 5% 5% & 10% 10% & 25% 25% & 50% 50%
2013 23.20% 13.40% 10.11% 7.31% 3.30%
2010 20.64% 11.98% 8.70% 6.01% 2.37%
2007 20.66% 12.92% 9.61% 7.27% 3.56%
2004 20.83% 12.53% 9.41% 7.27% 3.53%
2001 23.91% 15.20% 11.87% 9.20% 4.92%
1998 23.63% 14.79% 11.63% 9.12% 4.44%
1995 23.53% 14.46% 11.71% 9.43% 4.39%
1992 21.19% 13.99% 11.39% 9.42% 4.39%
1989 20.71% 13.93% 12.08% 9.77% 5.11%
1986 25.68% 15.99% 12.97% 10.48% 5.63%
1983 23.64% 15.54% 13.20% 10.76% 5.66%
1980 26.85% 17.13% 14.80% 11.91% 6.10%

 Progressive since ATR is higher at higher incomes

ATR of every income group listed above…
 …increased between 2010 and 2013
 …decreased between 2001 and 2010
 …significantly decreased between 1980 and 2013
Comparing taxation outcomes in 2013 to 1980, ATR of
_____ was _____ lower in 2013 than in 1980…
 … Top 5% ... 13.59% = (26.85–23.20)/26.85
 … 5% to 10% ... 21.77% = (17.13–13.40)/17.13
 … 10% to 25% ... 31.69%
 … 25% to 50% ... 38.62%
 … Bottom 50% ... 45.90%
149
 Measuring the Degree of Progressivity of a Tax:
 Can we reduce these observations on the distribution of
tax payments to a single number/measure?

Definition of the Stroup Coefficient of Tax Progressivity

Fraction of Total Income, Fraction of Total Taxes

Lorenz Curve
A
B Tax Concentration Curve
C

0 1
Fraction of Total Population

0
Stroup Coefficient – a measure of the degree of
progressivity of a tax, defined as the ratio of the area
between the Lorenz Curve and the Tax Concentration
Curve to the entire area below the Lorenz Curve
B
 as illustrated above, S= B +C
 for a proportional tax, the Tax Concentration Curve
would coincide with the Lorenz Curve => S=0
 for a progressive tax, the Tax Concentration Curve lies
below the Lorenz Curve => S>0
 value can range between 0 and 1 => larger value
reveals that the tax is “more progressive” (i.e., the
burden falls disproportionately more on high income
earners)
150
Measuring the Redistributive Capacity of a Tax:
 We already have a measure of income inequality (Gini
A
Coefficient => G= A+ B+C =2 A in previous graph)
 So, we could gauge redistributive capacity by observing
the change in Gini Coefficient brought about by the tax
 Let GI denote the “initial” (i.e., “pre-tax”) value of the
Gini Coefficient
 Let GF denote the “final” (i.e., “post-tax”) value of the
Gini Coefficient

Pechman-Okner Coefficient – a measure of the

redistributive capacity of a tax, defined as the percentage
decrease in the value of the Gini Coefficient brought about
by imposing the tax
G I −G F
PO=
 G I

 a proportional tax will not alter the value of G =>

GF =G I => PO=0
 a progressive tax reduces G => GF < GI =>
PO> 0
 value can range between 0 and 1 => larger value
reveals that the tax redistributes income to a greater
degree (i.e., the redistributive capacity of the tax is
greater)

151
Stroup Coefficient for the U.S. Federal Income Tax

1929 to 2010:
 high of .99796 in 1929; low of .44523 in 1969
 above .98 in every year up to 1939
 2008 = .70031; 2009 = .72834; 2010 = .72645 => only
three years since 1942 with value above .70
 Steady/consistent increase since achieving low in 1969
Pechman-Okner Coefficient for the U.S. Federal Income Tax

1929 to 2010:
 high of .14840 in 1943; low of .01492 in 1931
 below .05 in every year up to 1941

152
 2010 = .07899 (i.e., tax decreased value of G by
7.9%)
 stable since 1946 => between .07 & .09 in 48 of 65 years
Reconciling an apparent inconsistency…
 Stroup coefficient reveals that U.S. Income Tax was
highly progressive up through 1939
 Pechman-Okner coefficient reveals that redistributive
capacity of U.S. Income Tax was low before 1941

But if highly progressive taxes redistribute income, how

could these two outcomes simultaneously occur?
 Being “highly progressive” is a necessary – but not
sufficient – condition for a tax to redistribute income
 In order to significantly redistribute income and reduce
income inequality, a tax must be:
(i) highly progressive
(ii) substantial in size/scope

 Before the 1940s the tax was highly progressive, but

not substantial in size/scope
o % of adults on filed tax returns 12.1% in 1939
(below this level before then)
o Taxes paid as % of societal income 1.22% in 1939
(below 1.8% before then)
 In early 1940s the size/scope of the tax was greatly
increased (at the same time when degree of
progressivity was dramatically decreased)
o % of adults on filed tax returns 73.4% in 1944
(above 70% ever since)

153
 o Taxes paid as % of societal income 9.77% in 1944
(above 7% ever since)

 The “expansion of the tax to the masses” made it less

progressive, but it also allowed it to redistribute income

154