Edward Glaeser The Quarterly Journal of Economics 2005 Traviss Cassidy and Salama Freed June 14 2011 Outline of Discussion Introduction Basic Tenets of Hatred The Hatred Model Basic Sequence ID: 303690
Download Presentation The PPT/PDF document "The Political Economy of Hatred" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
The Political Economy of Hatred
Edward
Glaeser
The Quarterly Journal of Economics
2005
Traviss Cassidy and Salama Freed
June 14, 2011Slide2
Outline of Discussion
Introduction
Basic Tenets of Hatred
The
Hatred Model:
Basic Sequence
The Hatred Model: Expected Benefits
The Hatred Model: Stage 2
The Hatred Model: Stages 1 and 3
Examples:
Racism in the American South United States
Anti-Semitism in 19
th
Century Europe
Islamic Hatred of AmericansSlide3
Basic Tenets of Hatred
“People say they hate because the object of their hatred is evil.”(
Glaeser
, 2005)
Hatred “stems from seeing oneself under attack” (
Baumeister
, 1995)
Hatred is often a characteristic of an entire group, not just those who are actually guilty of the crimes from which the hatred stems.
Glaeser
argues in this paper that the spread of hatred is successful when voters have little incentive to investigate and dispute a false signal. Slide4
The Hatred Model
The Basic Sequence
Politicians spread hatred when their policies will harm a particular group in an effort to create desire to exclude or deprive the particular group.
There are four periods of the model:
Politician decide to spread hate filled message about particular group (“out-group”) to those not a member of that group (“in-group”).
In-group members receive signal, decide on a series of two actions
Whether the allegation deserves investigation, and
If credible, whether to protect themselves against out-group.
Members of both groups vote, and the winning politician gets their way
Out-group members may harm in-group members as a result of policies implementedSlide5
The Hatred Model
Expected Benefits
Three main constituents seek to maximize what is important to them:
Politicians maximize expected support minus the total cost of their campaign.
The in-group maximizes expected utility through all actions in the game where
Utility= Net Income +Expected harm from out-group – investigation costs – self-protection costs
The out-group maximizes their net income.
Note they are passive in the basic model (in that they are not receiving additional benefits) but are given an opportunity to become active in the future. Slide6
The Hatred Model
The Basic Sequence
Stage 2
Politician sends false signal to in-group
of harmful nature of out-group
Out-group is harmful with probability one or
θ
/(
θ
+
φ
(1-
θ
)) based on distribution
Out-group is harmless
In-group receives signal of out-group danger
In-group chooses to protect themselves
Do nothing
Investigate
Don’t Investigate
Out-group is harmfulSlide7
The Hatred Model
Stage 2: The Response to Potential Harm
After receiving a signal, the in-group:
Decides whether to spend
S
to investigate the signal, then
Chooses whether to spend
R
to protect themselves from potential harm or reduce damages from said harm (by a factor of
δ
<1)
Note the
benefit they will receive from protecting themselves outweighs the
amount spent for in-group to protect themselves, so self-protection is optimalProp 1: Investigating and learning the truth of a story is only optimal iff the cost of the search is less than the gains from the investigation.The cost of investigating the signal varies among the in-group, based on prior knowledge, frequent contact, or education, as well as other factors.
Less educated people will tend to dismiss false stories (and will likely have means to search for truth)People who have frequent contact with the out-group will dismiss false stories as well.Slide8
The Hatred Model
Stage 2: The Response to Potential Harm
The expected costs of the in-group protecting themselves are based on their decision to investigate:
If they choose not to investigate , expected cost is cost of self-protection plus reduced damage due to self protection (regardless of whether group is harmful).
If they choose to investigate:
If the story is false, expected cost from self-protection and damage is zero
If the story is true, expected cost is the cost of self-protection plus the reduced cost of damage after self-protection.
Total expected cost of protection is
Where
is the damage to the in-group’s net income,
is the probability an out-grou
p is harmful, and
is probability a politician will spread a false message.
Slide9
The Hatred Model
Stage 2: The Response to Potential Harm
When the out-group is harmful, politicians do nothing because the message will be transmitted without their additional help and at no cost to them.
The portion of the in-group who has chosen to investigate believes that the out-group is harmful with probability one.
The portion of the in-group who has not chosen to investigate believes that the out-group is harmful with probability
Note that the probability of success by politicians to spread negative messages adversely affects the in-group’s ability to discern a truly harmful group from a harmless one who has been negatively targeted.
If the out-group is harmless and the politician sends a false message:
A share who has chosen not to investigate believes the group is harmful with probability
.
The share who has chosen to investigate finds that the message is false and believes the group is harmful with probability zero.
We will focus on the spread of negative messages and hatred of a harmless group.
Slide10
The Hatred Model
Stage 1: The Supply of Hatred
Politicians can spread hatred among in-group members at cost
K
.
K
differs across politicians based on their abilities to effectively spread hate and the ability to distort the truth.
Voters do not know the value of
K
, but do know that
K
is drawn from a distribution.
Assuming two politicians with the fixed government policies of redistribution based on income:
Income of both the in-group and out-group is characterized by a continuous distribution
The mean income of the in-group minus mean income of the out-group is
, where
p
is the proportion of the population that belongs to the out-group
Slide11
The Hatred Model
Stage 1: The Supply of Hatred
Our two candidates propose a tax level
τ
on their income
People whose income is below the average receives a transfer in the amount of
.
The two candidates propose two different tax levels,
(
The candidate who is against redistribution proposes tax level
The candidate who is in favor of redistribution proposes tax level
Those who support the anti-redistribution candidate:
Receive a financial gain of
-
.
Effectively reduce the income of the out-group by a factor of
-
it is beneficial for those who hate the out-group to reduce that group’s resources.
Support the hatred spread against the poor out-group
Are generally the very rich.
Slide12
The Hatred Model
Stage 1: The Supply of Hatred
The pro-redistribution candidate is supported :
By the out-group or those within the in-group who believe the out-group is harmless only
iff
-
By haters only if the amount gained from redistribution less the cost of potential harm done by the out-group after redistribution is net positive.
Prop. 2: The pro-redistribution candidate benefits from hate
iff
, i.e. if the mean income of the in-group is less than that of the out-group.
How do we choose to support a candidate?
The very poor will always support the pro-redistribution candidate.
The very rich will always support the anti-redistribution candidate.
The actions of the middle-income group members will vary , as they will choose to support the anti-distribution candidate when they hate the out-group.
Slide13
The Hatred Model
Stage 1: The Supply of Hatred
Now assume
>0 (in-group wealthier than out-group)
Only the
anti-redistribution
candidate spreads hatred
Prop. 3: There is a maximum cost
K
* that the
anti-redistribution
candidate is willing to incur to spread hatred
K* is rising with
K* falls with the out-group’s proportion (
) of the electorate
K* is rising with (the non-stochastic component of)
s
Share of
anti-redistribution
candidates that spread hatred is falling with (the non-stochastic component of)
K
Slide14
The Hatred Model
Stage 1: The Supply of Hatred
= G(K*) in equilibrium
In-group voter’s subjective probability that the politician will send false signal is endogenous
Notice the opposing effects of increased information technology
Technology that lowers the cost of communication lowers
s
and
K
Lower
s
means more people will choose to investigate the veracity of a story
Hatred-creating stories less effective
Lower
K means it is easier for politicians to spread hatred-creating storiesLevel of hatred increases
Slide15
The Hatred Model
Stage
1
:
Antiminority
Policies
on out-group members
Support for the
anti-redistribution
candidate will rise with the level of hatred
iff
s/he is less generous to out-group:
Prop. 4: (Given that a mathematical condition [p
. 63]
holds)
K*
is rising with
and falling with
Anti-minority
policies increase supply of hatred when proposed by
anti-redistribution
candidate and decrease supply of hatred when proposed by
pro-redistribution
candidate.
What matters is whether anti-minority
policies increase or reduce the differences in how candidates treat out-groups
Slide16
The Hatred Model
Stage
1
: H8ing the H8erz
If
anti-redistribution
candidate spreads hatred, at cost
A
+
α
the
pro-redistribution
candidate can spread hatred against the
anti-redistribution
candidate
Out-group members and in-group members who searched will believe
anti-redistribution
candidate will impose costs of
ω
> 0. Prop. 5: There is a value α*
such that A + α* is the maximum amount the
pro-redistribution candidate is willing to pay to villianize the anti-redistribution candidate.
Given two mathematical conditions [p. 64], α* is falling and K* is rising with A
and , and
α* is rising with the proportion of voters who search
Slide17
The Hatred Model
Noneconomic Issues, Multiple Issue Elections, and Exclusionary Policies
:
non-income variable
)
: net tax payment
Support for the
pro-redistribution
candidate (p. 65)
The high tax candidate will support hatred if out-group is well endowed in
, and the low tax candidate will support hatred if out-group is poorly endowed in
.
All the previous results still apply
Slide18
Supporting Historical Examples
Anti-Semitism in 19
th
-Century Europe
Anti-Americanism in the Muslim world