Lecture 9

Document Sample
Lecture 9 Powered By Docstoc
					Psychology of Poverty

   Sendhil Mullainathan
 Design of Redistribution Programs

 Specific Psych of Poverty: Temptation tax

 Specific Psych of Poverty: Packing Problem
 Design of Redistribution Programs

 Specific Psych of Poverty: Temptation tax

 Specific Psych of Poverty: Packing Problem
Transfer policy: Behavioral lens
 Delivery
      How do individuals engage participation decisions?
      E.g., defaults and 401(k)s
 Design
      What problems do benefits leave in hands of
      E.g., food stamps and planning problem
 Disincentives
      How do individuals respond to program incentives?
      E.g., TANF time limits and commitment
Transfer policy: Design
 Benefits need to take a form that is useful to individuals
 Rational model:
      Cash is best, people can smooth benefits over
       time, etc.
      PDV of benefit really all that matters
 Behavioral model:
      Complexity and planning
      Smoothing and self-control
Transfer policy: Meeting Needs

 Food stamp example (Shapiro 2005)
       Want benefits to translate to improved nutrition
       Big to small problem, limits to self-control
       Declining nutrition over month
 ―Give it to us in two installments. At the end of the month I'm
   dying [for money]. If you got it on the 1st and 15th, or
   whatever, it would be so much better. Checks or coupons, it
   doesn't matter, either way, but it does not last a month. The
   second part of the month is always a struggle.‖
An example: meeting preferences

 •   Timing of EITC
     • Lump sum vs. advanced
     • Want income to translate to consumption
     • Commitment device?
 Provide benefits without distorting incentives (or even while
   reinforcing desirable incentives)
       Work incentives a particular concern
 Rational model:
       Programs like TANF create a moral hazard
       Government interests, recipient interests at
       Create restrictions, e.g., time limits
 Behavioral model:
       Program incentives mediated by psychology
       Incentives interact with, e.g., self-control
       Recipients act on incentives as construed
False Fear of Moral Hazard
 Fear of moral hazard inhibits redistribution
 Suppose however that b < c for most b but β(b) <
  c for behavioral types
      Driving without a helmet
      Teen pregancy
 Those who engage in behavior are behavioral
    Transfer of min(b)-c would induce no moral hazard
 Note: existence of those doing it leads to
  misestimation of f(b) type distribution under
  assumption of rationality
    Construal of Incentives
 Eissa and Hoynes (2004): “A consistent and somewhat
  puzzling finding in the empirical literature on the EITC and
  labor supply is the large response of the participation
  decision and the lack of any response by the reported hours
  worked by taxpayers in the labor force.”
 Some evidence that people do not understand marginal
  incentives: information “induce[s] treated clients to increase
  their EITC refunds by choosing an earnings level closer to
  the peak of the EITC schedule. Clients treated by complying
  tax professionals are 10% less likely to have very low
  incomes than control group clients.” Chetty and Saez (2009)
 Interesting possibility:
    Assume can still work and get benefits, but do not
      understand marginal and therefore hours substitution
    Behavioral tagging: can transfer to these types without
      affecting (hours) incentives
Behavioral Tagging
 Direct implication of above
      Ironically bad behaviors present redistribution
       opportunities with little cost
           Must still worry about elasticity of behavioral type
           Note: problem is simpler if behavioral type were
            known to be correlated with b
           Then these behaviors themselves are a poverty tag
               Note contrast between arbitrage for welfare benefits vs.
Behavioral Tagging
 UI:
   Transfers to deal with loss aversion in job search
   Threat of Re-employment services
           Black et. al. estimate that the program reduces mean weeks
            of UI benefit receipt by about 2.2 weeks, reduces mean UI
            benefits received by about $143, and increases subsequent
            earnings by over $1,050. Most of the effect results from
            asharp increase in early UI exits in the treatment group
            relative to the control group.
      More fine-grain incentives for inputs (resumes sent)
       rather than outputs
 Key insight: incentives can match the behavioral type
Benefit of Behavioral Tagging
 In some cases:
      Transfer without fear of moral hazard
      Low types are behavioral
 In other cases:
      Undo what appears to be “moral hazard” without
       giving up insurance
      High types are behavioral
    Incentives and Self-Control

 TANF time limits
      Serve as a commitment device (Fang and
       Silverman 2004)
      Interests of policymakers and recipients align
 Moral hazard confounded with behavioral tendencies
      Moral hazard is not the only mechanism for
      Benefits can induce, indulge procrastination
      Some behaviors are more likely to be evidence of
       behavioral tendencies than moral hazard (e.g.,
       teen pregnancy)
Transfer policy: Design
 Need more research on demonstration projects and
  evidence on
      Cash transfers or structured transfers?
      Moral hazard or procrastination?
 Understanding could lead to transfer programs with
  much less moral hazard.
      Potential for more efficient transfers.
 Need more research, evidence, data
 Design of Redistribution Programs

 Specific Psych of Poverty: Temptation tax

 Specific Psych of Poverty: Packing Problem
Aside: Two Ways to Apply
 Same psychology we always apply

 A different ―twist‖ on the same psychology
Vegetable Vendors – Stark Example
 Simple production function
     Purchase fruit in the early morning
     Sell through day
 Basic working capital needs
Fruit Vendor
 Simple production function
     Purchase fruit in the early morning
     Sell through day
 Basic working capital needs
Table 1-Business Characteristics of sample population
Detail                        Percentage of Average amount   Profits     per
                              respondents      purchased*    day*
1. One trip a day to the 89.7%                 Rs. 1075.3    Rs.110.5
market- normal days                            (589.2)       (54.7)

2. twice or more trips a day( 8 %        Rs.707.5            Rs.95.6
total amount purchased per               (422.6)             (46.1)

3. once in two days trip to 2.3%         Rs. 1034.8          Rs.97.2
the     market      (amount              (515.8)             (44.3)
purchased per trip)
4. good days a week         98.9%        Rs. 1666.3          Rs. 186.6
                                         (834.3)             (83.4)
5. festival days             91.5%       Rs. 2580.7          Rs. 318.2
                                         (1543.7)            (187.3)
Table 3- Meter loans for financing
1. % of sample size that takes daily loans                                  69.4%
2. % of sample size that takes daily loans for more than 15 days a month    65.7%

3. average number of days in a month that respondent takes a daily loan for 25.8 days
    working capital

4. average number of years of taking daily loans                            9.5 years
5.average daily interest rate                                               4.9%
6. % of total meter loan borrowers who borrow from the same moneylender 67.7%
7. Average of maximum that can be borrowed as a daily loan                  Rs. 4098.6

8. % of meter loan borrowers who feel there is no other way of doing 63.8%
   business and the interest is unavoidable
 Persistent borrowers
 At very high rates
 Stark implication:
     One less cup of tea a day.
     In 30 days will have doubled income.
 Significant foregone income
 Is this a rationale for consumer protection?
     Tempted to debt?
Related Puzzles
 The poor borrow at very high rates
 Poor appear to have very high return
  investments which are divisible
      Fertilizer (Duflo, Johnson, Kremer)
      Capital drops diMel, McKenzie and Woodruff
 Paying off high interest rate debt is a special
 Generally the poor appear more myopics
Implications of high interest rate
                          u ' (ct )  R u ' (ct 1 )
 Discount future heavily (δ low) or
 Future marginal utility large relative to today
       Consumption growth large
            u’(ct+1) low so ct+1 high
            Note: this is stronger than saying that marginal product
             of capital is high.
                Some existing studies suggest this as well.
       Particularly sensible for transitory shocks (e.g. health).
            But examples span even working capital uses (e.g. crop
Understanding Poverty
 To fit these facts current models must assume

                Poor are very myopic
         Poor are quickly becoming non-poor

 Since the latter appears counter-factual, we are
  forced to conclude the former
How to incorporate psychological
 Time inconsistency obviously helpful in resolving
  myopia tension
      Standard model conceptualizes it as δ low
 Present bias literature suggests more nuanced view
    Individuals have inconsistent time preferences
    Can be both myopic and farsighted

 Two ways to incorporate time inconsistency
    Constant time-inconsistency
    Inconsistency problems which vary with income

 Need simple framework to do both
Model of Self-Control
 Time inconsistency often conveniently measured in
 But fundamentally it must be about consumption
      Tomorrow’s self wants to consume more than today’s
       self wants him to
      Tomorrow’s self wants to consumer more X than
       today’s self wants him to
          Cigarette smoking

          Sleep in the clocky example

          Unhealthy foods
Choosing Movies             Read, Loewenstein & Kalyanaraman (1999)

 Subjects given opportunity to choose a movie
  video from a set of 24 titles
     Four Weddings and a Funeral
     Schindler’s List
 When choosing for today: 56% choose low-
 When choosing for next Monday, 37%
  choose low-brow
 When choosing for second Monday, 29%
  choose low-brow
Temptation goods
 Small modeling deviation:
   Temptation not just about how much to
    consume today but what to consume today
   Easier to model how temptation changes with
    income if we explicitly allow for temptation
        Can talk about whether temptation goods are
         non-normal goods

 Will assume ―sophistication‖ in consumption
   Not essential to our analysis
Set Up
 Two periods in most examples
 Two types of index goods: x and z
    x consumption: no time inconsistency
    z consumption: only present selves like it
    Instantaneous utility in each period u(x) + v(z)
 Period 1’s decision utility:

                 u(x1)  v(z1)  u(x2 )
        Income each period yt and initial wealth w0
        Production function f(). Sometimes for simplicity will
         just assume rate of return R
 Will assume ―sophistication‖ in consumption

Aside: Modeling Temptation
 We can think of this in a more general way.
 Let c be a vector of consumption goods
    Two utility functions uF(c) and uT(c) be forward looking and
     temptation utilities. So individuals maximize

                 v        F v        2 F v
             u ( define indices   u z(c) and
     Result: Can c1)  u (c 2 ) x(c) and(c 3 ) L utility
     functions u(x) and v(z) so that the individual consumption
     maximization is represented by the maximization over how
     much to spend on x and z.

                v          v              v
          u(x(c1))can vary)) u(x(cconsumption
 Key observation
    Goods’ temptation
                       v(z(c1 with level of 2 ))
 Two periods in most examples
 Two types of index goods: x and z
    x consumption: no time inconsistency
    z consumption: only present selves like it
 Instantaneous utility in each period u(x) + v(z)
 Period 1’s decision utility:
            u ( x )  v( z )  u ( x )
                 1        1           2

 Income each period yt and initial wealth w0
 Production function f(). Sometimes for simplicity will
  just assume rate of return R
Generalized Euler Equation

 Traditional Euler Equation:

            't) 
            ( ' u
             c ( (
             ) w   c
            u f t 't)

 Generalized Euler Equation

   ') 
'   '1
u( t1
c t( (
 )w c
    '1 t
    c 
 t   )[ )
 Temptation tax:
    Every dollar transferred into the future is ―taxed‖ by
     temptations; future selves will waste some of it.
Poverty and Myopia
 Two forms of ―myopia‖: δ and z’(w)
 Original puzzle
    Third explanation: myopia in the form of high z’(w).
 Why is this different?
    Because z’(w) can vary systematically with w
    Individuals can control the value of z’(w) they face and
     hence the tax.
    All our results come from this.
 These considerations that traditional estimations of
  discount rates be biased
     Relation to Hyperbolic Model
      Special case in more general model where

                     u (c)  u (c)
                       T           F

      So a special case is

           u(x)  u(c) and v(z)  (1 )u(c)
                   ˜                     ˜
      In the more specific formulation

                          x1                z1
                u(x t )  t        v(zt )  A t
                       1                 1 
The shape of temptation
 Two important cases:
     z’(c) constant (Non Declining temptation)
          Rich and poor face similar time inconsistency
          Includes case of z’(c) = 0
     z’(c) declining
          Rich face less time inconsistency problems
Example Analysis - Indonesia

                           Bjorkegren (2009)
Example Applications
 Attributions of impatience
 Impact of future income
 Poverty trap
 Response to uncertainty
 Investment features
 Role of credit
 Money Lender
 Testing this model
 Design of Redistribution Programs

 Specific Psych of Poverty: Temptation tax

 Specific Psych of Poverty: Packing
Understanding Poverty

  Current theories/approaches to behavior in
      Education
      Health
      Crime, etc.
  Really about covariates of poverty
  Is there something intrinsic to poverty?
  Our approach: Scarcity
      Conditions of scarcity (in money, time…) produce their own
      This psychology, interacted with scarcity, produces specific
The Packing Problem:
A Suitcase metaphor

2 people packing
     One a large suitcase; the other a small suitcase
     Larger suitcase - packs everything with room to spare,
      or at least all important things, then chooses among
      remaining, less important items
     Smaller suitcase - tries to pack the essentials, and
      might need to choose among remaining
      important items
     The larger suitcase allows you to leave it
      partly unpacked. In case something else
      arises. The smaller suitcase does not..

 (―The Packing Problem‖ in Complexity /
                      Computational theory)
The Suitcase Metaphor

 Slack makes it easier to pack
   Slack reduces cognitive costs: easier to pack and travel with partly empty suitcase
   than a full one.

 Bigger suitcase means slack is ―cheaper‖
   Slack is easier to maintain when suitcase is larger because what is being given up
   to buy slack is cheaper.

 Problem complexity higher when suitcase is small
   Larger suitcase means easier problem, requires less vigilance, both re decision
   quality or resistance to temptation; hence, provides greater margin for error.
   Smaller suitcase presents more complicated packing problem.

 A consequence: error
   Consequences of mishaps (misfortunes & mistakes) greater without slack.
   Especially important ―mistake:‖ giving into temptation. The temptation looks like a
   graver error.

 Lack of slack → The large suitcase person (typically greater slack) will
   appear more competent, a better packer.
Big suitcase can also be tightly packed

Poor in Trenton??!!

   The poor in Chennai are packing summer clothes in
    a small suitcase; the poor in Trenton are packing
    winter clothes in a (somewhat - but not enough)
    larger suitcase:

   Poor on wall street…
   (& rich in Chennai..)
Psychology of Scarcity

 Greater attentiveness (& knowledge...)
 Forced to make tradeoffs
 Global bracketing
 Objective -- as opposed to relative -- valuation
 Horizon shortening
 Ego depletion
 ….
Taxi fare when you first get in?…




           Low SES   High SES
Imagine that a friend goes to buy an appliance
priced at $100($500/$1000). Although the
store’s prices are good, the clerk informs your
friend that a store 45 minutes away offers the
same item on sale for $50 less. Would you
advise your friend to travel to the other store to
save $50 on the $100($500/$1000) expense?

                            (Crystal Hall dissertation)
Percent traveling to save $50:

                              $100   $500 $1000
HI    (Princeton commuters)    54     39   17
LI   (soup kitchen)            76     73   87

       Willingness to travel to save $25…


                                                                            Low SES
                                                                            High SES



        High reference point ($800/$775)   Low reference point ($100/$75)
Suitcase (Thinking about tradeoffs)

 Ted and James are on vacation. Ted is traveling
 with a small and very full suitcase. James’ suitcase
 is larger and has some space. They visit a store
 offering running shoes on sale.
    James asks ―Do I like these shoes?‖
    Ted wonders ―Do I like these shoes?
     What, of at least equal size, do I take out of the
     suitcase if I buy them – and is it worth losing?!‖...
% who think about what they might not buy instead…


                                      Low SES
                                      High SES



               TV          Toaster
% who think about what they might not buy
instead… (India)



                                       Low SES
                                       High SES


                 TV            Mixie
    Scarcity in Time (tradeoffs)

 George is a tourist visiting Rome for a weekend; Giorgio is in
  Rome for 6 months. Sunday afternoon, they both walk by a
  lively street fair.

 Giorgio wonders, ―Am I in the mood to listen to performers?‖

 George thinks, ―Am I in the mood to listen to performers?
  What in my long to-do list will I forego in return…?
1) Scarcity implies broader frame
  Even ―positive expenditures‖ are tainted with tradeoffs
      Spending money on a basic good for a poor person
       turns it into ―luxury‖
      Spending time with family for a busy person turns,
       e.g., family time into ―luxury‖?…

2) Second piece of puzzle
   Lack of slack means giving into temptations more
Temptation (& its discontents)

 Abe and Ben walk by a clothing store. They see
  an attractive leather jacket and in a moment of
  weakness both make an ill-advised purchase.

 Abe goes home and thinks ―What a bad purchase‖

 Ben goes home and thinks ―What a bad purchase.
  Now I won’t have the money to repair my car. That
  means I may be late for work, which might lead
  me to…‖
Who appears more myopic?…
(Note: could rationalize the attribution, but such attribution may still
produce a bias…)
Temptation (& its discontents)

  Abe and Ben pass by a pastry shop and see a beautiful
   eclaire. Both have been trying to eat healthier.
  Abe finds mental reserve and resists the temptation.

  Ben, whose mind is occupied with the results of the
   upcoming bar exam succumbs and buys the pastry.
  Charlie is stressed about whether he’ll be able to find
   enough money next month for his child’s day care. He also
Ego Depletion

 Executive control requires energy
 Like a muscle
      Fatigued with use
 Fatigued through work / distractions
 (commonality of executive control across domains)
 Stressors/ego depletion may make it harder to exercise
 Vegetable vendors
      Little temptations (cup of tea, dosa) are even harder to resist
      ―Oasis of pleasure in the midst of misery‖
How do we measure depletion?

                        (Sugar cane Harvest
                         in India)

Kosslyn’s miniCog

Ego Depletion
Racists get depleted through encounters with Blacks
       …Participants were led to a different room, where a second experimenter (E2)
        was waiting for them. They were told that there would be a delay before the
        second cognitive task, and they were asked to help with the creation of
        stimulus materials for a different experiment. For half of the participants, E2
        was White, and for the other half, E2 was Black.
Decision quality
 Anne and Beatrice are offered a financial
 Anne reads it over carefully and notices that
  there are fees which make the advertised
  interest rate look bad.
 Beatrice’s mind is on her upcoming rent
  payment, school fees, medical expenses —
  ‖How will she make ends meet?!‖ She signs
  the contract.
    Money (the poor) vs. Time (you)
Tradeoffs           What will I not buy?    What will I not do?

Temptation           Shouldn’t have        Shouldn’t have
                     bought                committed

Error/Distraction      Forgot to pay       Forgot to do, or pay…
                                           (esp. w.out reminders..)

Bad borrowing       Given what you owe,      Given what you owe,
                    what are you doing       what are you doing
                    spending?!               schmoozing?!..
   Irony of Poverty

 Poor must make higher quality decisions
   Packing problems are harder
   Must resist temptations more
   Can’t afford mistakes

 Poor are in worse position to make higher quality
     Distracted by other stressors/decisions/conditions
     Depleted by challenges/temptations/past failures
     Unappreciated! (Both they and their packing problems...)