Fairness, Incentives, and Salience in the Demand for Redistribution

Download Report

Transcript Fairness, Incentives, and Salience in the Demand for Redistribution

Fairness, Incentives, and Salience
in the Demand for Redistribution
Christina Fong
Department of Social and Decision
Sciences
Alexis de Tocqueville Democracy in America (1835), Book II, chapter VII
The Americans...are fond of explaining almost all the actions of their
lives by the principle of self interest rightly understood; they show with
complacency how an enlightened regard for themselves constantly
prompts them to assist one another and inclines them willingly to
sacrifice a portion of their time and property to the welfare of the state.
In this respect I think they frequently fail to do themselves justice; in
the United States as well as elsewhere people are sometimes seen to
give way to those disinterested and spontaneous impulses that are
natural to man; but the Americans seldom admit that they yield to
emotions of this kind; they are more anxious to do honor to their
philosophy than to themselves.
Today’s Agenda
• Overview research on behavioral motives for
public redistribution.
• Research is located at intersection of several
enormous literatures on causes and consequences
of social policy
– Political history
– Public economics
– Public opinion, political psychology, behavioral
economics
Motivation: Step Toward a
Larger Goal
• Explain dramatic differences in the quantities of
transfers across redistributive situations.
– In 1995: Sweden spent 33% of GDP on social
expenditures. U.S spent under 16% .
– Foreign aid expenditures roughly 10x greater in
Northern Europe than in U.S.
– Dramatic and uneven growth of OECD public sectors
since WWII.
• Sweden: General outlays 31% of GDP in 1960, 64.5% in 1985
• US: General outlays 27% of GDP in 1960, 36.7% in 1985.
– 24% support spending on “welfare,” 64% support
“assistance for the poor.”
– Charitable donations raised after Sept. 11.
Two Main Questions
• Why and to what extent do people care
about others in various settings?
• Why and to what extent do people place the
responsibility of social welfare in the hands
of the public sector (governments) rather
than the private sector (markets,
communities, families)
Traditional Econ View: #1
• Selfish median voter
– Median Y less than mean
– Even if not, risk aversion creates insurance
motives
• Abundant, striking evidence against pure
self-interest
• Consequence: much effort on the wrong
questions
Traditional Econ View: #2
• Simple altruism and variants
– A positive weight is placed on recipient’s utility from
consumption, or utility from one’s own gift or transfer.
• Open question: to what extent does behavioral
evidence contradict this?
– Generosity may depend on perceptions of the poor,
group membership, social distance
– Generosity may vary across situations, cultures, time
• If generosity is conditional and situation variant,
traditional theories lose their power unless we can
find simple, empirically & historically disciplined
assumptions about:
– Behavior
– Institutional environment
Method
• Establish a set of empirical regularities using
sensitivity analysis on different types of data
• Begin with traditional model
U¹=(1-α)u(c2)+αu(c2)
– Traditional assumption is that transfers made only from
rich to poor, so this is a form of inequality aversion
• U¹ =u(cr)-α[u(cr)-u(cp)]
• Attempt to incorporate reasonable behavioral
assumptions into this framework
– Formulation of α.
– Attention to beliefs, preferences, incentives, and
constraints.
• Take history into account when deciding which
variables to endogenize
Published Behavioral Effects
• Americans support less redistribution if they think
poor are lazy rather than industrious but unlucky
(Kluegel and Smith 1986)
• Socioeconomic characteristics have surprisingly
small effects compared to “lazy poor” effects and
“lazy poor” effect is not spurious (Fong 2001).
• Effect is widespread, occurs in several European
countries and in different contexts (Bowles, Fong,
Gintis Forthcoming).
• Americans exhibit racial group loyalty in their
attitudes to redistribution (Luttmer 2001)
• Americans exhibit negative exposure effects if
exposed to recipients with undesirable traits
(Luttmer 2001)
“Lazy poor” effects are enormous
Figure 1.
Estimated Probabilities for Four Categories of People
Full Sample (N = 2738)
0.6
Estimated Probabilitiy
0.5
0.4
Believes in Luck
Believes in Effort
0.3
Low Socio-Econ Status
High Socio-Econ Status
0.2
0.1
0
1
2
3
4
Support for Redistribution
5
6
Recent Findings
• Substantial amount of generosity that is not
conditioned on beliefs about causes of income,
group membership, or social proximity.
– Substantial and VERY robust correlation between
domestic and foreign public generosity
• In country-level expenditures
• In individual-level attitudinal support
• Individual sense of moral duty to help poor countries predicts
country level expenditures on domestic and foreign transfers
• Consistent with research on cultural traits
(Hofstede 2001)
Country-level Expenditure Data
Dynamic models predicting Social Security Transfers (%GDP)
Independent Var.
Naive LSDV
First Difference
FD2SLS
Coeff.
Coeff.
Coeff.
(S.E.)
0.108**
(0.042)
(S.E.)
IVfor SSTRAN lag
ODA (%GDP)
6.295***
(0.678)
ODA (%GDP) lag
(S.E.)
0.791
(0.811) 0.760
(0.828)
3.239***
(0.772) 3.398***
(0.786)
ODA (%GDP) lag2
Real GDPpc (1000s) -0.425*** (0.122)
-1.221*** (0.155) -1.172*** (0.158)
Unemployment rate 32.915*** (3.982)
19.597*** (4.687) 23.716*** (5.001)
% pop. over 65
83.689*** (12.337)
Pop. (millions)
0.029
Open
-0.038*** (0.009)
(0.017)
Open lag
-0.009
(0.009) -0.009
(0.009)
-0.000
(0.008) -0.004
(0.009)
Time
0.152***
(0.040)
R2
.9460
.3814
.3435
N
441
441
394

Individual-level Attitudinal Data
Country fixed effects regressions predicting support for domestic redistribution using 1989 Eurobarometer data
(Dependent variable is standardized with a mean of zero and standard deviation of one.)
Independent Var.
Coeff
(S.E.)
Coeff.
(S.E.)
Helping poor countries of little importance (D) 0.267**
(0.110) 0.239**
(0.108)
Helping poor countries of important (D)
0.356***
(0.105) 0.287***
(0.104)
Helping poor countries very important (D)
0.455***
(0.105) 0.374***
(0.104)
Poor are lazy (D)
-0.428*** (0.043)
Predicted exposure to poverty (standardized)
0.153***
(0.039)
Pred. exposure*poor are lazy
0.342***
(0.046)
2nd Income quartile (D)
-0.095**
3rd income quartile (D)
-0.188*** (0.045) -0.161*** (0.043)
4th income quartile (D)
-0.207*** (0.048) -0.190*** (0.046)
Age
-0.007*** (0.001) -0.006*** (0.001)
Age when education ended
0.002
(0.005) -0.004
(0.005)
Male (D.)
-0.000
(0.031) 0.012
(0.030)
2
(0.044) -0.073*
R
0.1093
0.1645
N (sample weights used)
7169
7169
(0.044)

Substitution between public and
private giving?
• Incentive for public giving: If transfers are private,
people may free-ride on each other's altruism.
Under-provision of transfers.
• Incentive for private giving: Taxes and public
transfers involve incentive costs
• Survey results: People who have given to charity
want more public redistribution but are more
likely to live in countries that in fact spend less.
• More work on this in progress with Jörgen Weibull
Merged Expenditure and
Attitudinal Data
Ordered Probit regressions predicting real ODA expenditures
Coeff.
S.E.
Has given to charity
-0.253*** (0.028)
Moral duty to help 2
0.075
(0.064)
Moral duty to help 3
0.291***
(0.055)
Moral duty to help 4
0.355***
(0.058)
Age
0.004***
(0.001)
Male
-0.030
(0.028)
Age when education ended 0.055***
N

(0.005)
7861
Note: Including income quartiles has no noticeable 
effect on the reported results but results in the loss 
of over 1300 observations.

Possible Modeling Approach
• α=f(C, A(êj),G(| i- j|+|êi- êj|)).
• ∂α/∂C>0, ∂α/∂A>0, ∂α/∂G<0.
• Can impose a shape consistent with idea
that decision rules may change suddenly
with the context (Loewenstein 2001)
– α might be sigmoid shaped function of the
arguments
Issues!
• Where do beliefs about effort come from?
• Are characteristics/work activities of the poor
endogenous?
• When does group psychology take effect?
• Are “lazy poor” effects as universal as they seem?
– Surprising result: NO “lazy poor” effect in some
countries!
– It is still not completely clear how to interpret the
effect.
• Is it a question of salience?
Salience
(What Follows is Work in
Progress!)
• Assumption: A individual’s attention is
focused on a variable and the variable
becomes salient when that person perceives
a consequence to different values that the
variable may take.
Stylized Fact
• Countries in which social expenditures are
either in large part work promoting or in
small part means tested have smaller “lazy
poor” effects.
OECD Social Expenditures Data
•
•
•
•
•
•
•
•
•
•
•
•
•
Old age pensions
Disability pensions
Occupational Injury and Disease
Sickness Benefits
Services for the Elderly and Disabled People
Survivors
Family cash benefits
Family services
Active labor market programs
Unemployment
Health
Housing
Other contingencies
Interaction Effect
• In Eurobarometer data, effect of belief that
poverty is caused by laziness on opposition
to redistribution decreases as social policy
of a nation becomes more work promoting
or less means tested.
• Data are severely limited
• What can theory say?
Optimal Redistribution with
Endog. Search Effort and Exog.
Work Requirements
• People may either work and earn income in
private sector or take government transfer
and meet certain work obligations of the
social policy.
• Probability of not getting private sector job
depends on luck and effort.
• Rich care about the poor
Two Steps
• Individuals choose job search effort given
the tax/transfer rate and exogenously
enforceable effort levels in social program
and exogenous work norms in private sector
alternative
• Taking the optimal effort function into
account, individuals choose their preferred
level of redistribution
Result
• Optimal redistribution increases in
exogenously enforced effort.
• The effect of effort expended on optimal
redistribution decreases as the exogenously
enforceable effort increases.
Interpretation
• Incentive problems can focus attention on
the disutility of effort and the labor market
activities of the poor and make this a salient
issue in redistributive politics.
What’s Next?
• Now that I know what to look for, I will
– Try to refine my characterization of welfare
states according to the incentives they provide.
– Test for interactions between subjective
concerns about incentives and “lazy poor”
effects.
• I will then pull incentives, salience, and
fairness together into one model.
Summary
•
•
In economics, trusted empirical regularities are
often a constraining factor.
Most of the progress in my research area has
been empirical. Four important effects:
1.
2.
3.
4.
“Lazy poor” effects
Unconditional generosity
Racial group loyalty
Social distance/proximity
Summary, Con’t
• However, I have found that these effects are not
universal.
• Under what conditions might the variables be
salient and have effects on redistributive
demands?
• I model incentive costs as the main consequence
of “laziness” that tax payers may focus on in their
decision over their optimal redistribution.
• This application area illustrates important
functions of theory: to illuminate the variables and
how they matter, and guide empirical
investigation.