Principles and practicalities in measuring child poverty Miles Corak Measuring child benefits: measu - PowerPoint PPT Presentation


PPT – Principles and practicalities in measuring child poverty Miles Corak Measuring child benefits: measu PowerPoint presentation | free to download - id: 133659-MTFmY


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation

Principles and practicalities in measuring child poverty Miles Corak Measuring child benefits: measu


Principles and practicalities in measuring child poverty - Miles Corak ... most low income households is very poor & families may be poor &/or alienated ... – PowerPoint PPT presentation

Number of Views:101
Avg rating:3.0/5.0
Slides: 32
Provided by: larso5
Learn more at:


Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Principles and practicalities in measuring child poverty Miles Corak Measuring child benefits: measu

Principles and practicalities in measuring child
poverty - Miles CorakMeasuring child benefits
measuring child poverty - Michael Mendelson Sid
  • Lars Osberg
  • Economics Department
  • Dalhousie University
  • Low Income, Poverty, and Deprivation
  • A Workshop sponsored and hosted by
  • Statistics Canada and the University of Toronto
  • June 5/6 2006

Different Focus Agreements, Disagreements
  • Agreements
  • Poverty is relative to social norms of
  • Focus
  • Identify Poor Administration or Evaluation ?
  • Type I Type II errors of classification
  • Disagreements
  • Market Basket or of Median Income ?
  • Word Picture to motivate criterion ?
  • Summarization rate gap ?
  • Omissions
  • Time Risk, Human Rights Link

Poverty economic deprivation - want of the
necessities of life
  • not only those things which nature, but those
    things which the established rules of decency
    have rendered necessary to the lowest rank of
  • Adam Smith (1776)
  • poverty is an absolute notion in the space of
    capabilities but very often it will take a
    relative form in the space of commodities or
  • Amartya Sen (1983)
  • MM MC agree

Capabilitiest Fit (commoditiest)
  • Poor - If Capabilities lt Minimum Necessary
  • Social Standards of time gt Minimum _at_ time t
  • Fit F (household size composition, location,
    health, disability, life skills, etc.)
  • Commodities alternative measures
  • Actual consumption or potential consumption
    (income) or specific necessities
  • Minimum Capabilities / Deprivation ??
  • Ambiguity of citizenship

Issue - to identify the poor or the number
deprivation of the poor ?
  • MM What is an adequate maximum child benefit
    for families with low incomes?
  • MMs issue administration of a particular
    program Canada Child Tax Benefit - implies
    specific information set (Xi)
  • Type I and Type II errors both matter
  • Good Predictor of need at individual level
    needed for effectiveness
  • Other Programs supposed to take care of
    special needs, adults

MC aim to define measure child poverty,
compare internationally define credible poverty
targets for society
  • Issue given all programs, macro cycle,
    demography, etc., how big is the problem?
  • Rate, average gap inequality of gaps
  • Note If Type I error Type II error, total is
  • Identification of deprivation of specific person
    not the issue
  • Statistical agency information set ?
    Administrative sets
  • Different criteria for success
  • MC objective to Minimize Type I Type II
  • MM objective to Minimize Type I Type II
  • MM Much discussion of classification errors

Practical Issues in Poverty Measurement
  • MM MC agree on dimensionality
  • accounting period year,
  • recipient unit family / household
  • measure after tax transfer money income
  • supplementation by specific item deprivation
  • MM MC disagree on poverty line
  • Relative Poverty or Budget based / Absolute
  • Implications for Initial Level and for Updating
  • LO Missing Issues
  • Summarization
  • Spell Durations
  • Human Rights link

MM1 Market Basket Budget
  • MM all relevant measures of low income or
    poverty are relative to contemporary living
  • Historical evidence - absolute poverty lines
    market basket budgets increase over time with
    average incomes
  • Non-transparent but appearance of objectivity?
  • Details of construction buried deep
  • wide range e.g. 18,342 to 30,402 for 4 in
    Toronto 1992
  • Major political plus iff source is credible
  • Major minus if change is suspiciously timed
  • Implicit link to norms, occasional, idiosyncratic
  • Not directly comparable across nations, time

MM2 reverse regression ???
  • MM proposes regressing the sample of families,
    with income as the dependent variable and the
    deprivation indicators as the independent
    variables, for families of different structures
  • Yi a0 a1D1 a2D2 ß1X1 ß2X2 ei
  • But complains With current techniques, we know
    only that renters have lower incomes than owners
  • Better estimate Prob(D1,D2 Yi ,X1,, X2 )
  • BUT no escape from arbitrary choice of minimum,
    necessary Probability of (D1,D2)

MC Explicitly Relative Poverty Line
  • Ho Established rules of decency change as
    society changes
  • Transparent link to current distribution of
  • LIS standard ½ median equivalent income
  • European Union 60 median equivalent income
  • Explicitly transparent, standardized, comparable
    over time across countries
  • huge plus for policy evaluation
  • Two issues initial level updating
  • Choice ½ or 60 - ?
  • Rapid change in median can imply
    counter-intuitive changes
  • Update for price changes OR Update for median
    real income changes ?
  • Moving average can diminish volatility

When does the choice of updating method matter ?
  • Polling data since 1946 - social norms track
    median earnings
  • IF median real incomes change, updating method
    matters sometimes has been historically
  • Canada 1960s, 1970s, USA in late 1990s, Ireland
  • BUT since 1981, not much change in median real
    incomes in Canada

Absolute or Relative Poverty Line ? Not much
change in real median incomes !
How best to motivate choice of initial level of
poverty line ?
  • MM general word picture of standard of living
    at poverty line should drive number
  • How to change specific items over time?
  • How to defend general criteria as non-arbitrary?
  • MC ratio of potential consumption to median
    household income implies number
  • less than half deprivation
  • LO ½ has intuitive meaning
  • 50/50 equity norm ( solution to ultimatum
  • Individuals always apply Reality check
  • number drives examples, which drive credibility
  • Comparisons drive Concern
  • Over time is poverty getting worse ?
  • Across countries do other nations do better ?

Item Deprivations An Increasingly Important
Complement to povline
  • Communicable to Public
  • Human Rights Perspective gt specific goods
  • Convention on the Rights of the Child - includes
    health and nutrition, clothing, housing, and
    other goods, services and opportunities necessary
    for normal physical, mental and social
  • Can see behind within family allocation of
  • to children, elderly, women.
  • Legal recognition of social norm of minimum
  • literature now emphasizes multi-dimensionality
  • Sens Capabilities Perspective
  • Social Exclusion Agenda
  • Social Wage or Cash purchase ? who cares?
  • Comparable across differing public-private policy

LICO orphan methodology
  • Relatively little discretionary income
  • LICO based on 20 E (food clothing house
    share of expenditure income, family size,
  • Year to year updating on CPI as per absolute
  • Rebased occasionally to community norms as per
  • Long time series useful !!!
  • for research politics
  • BIG MISTAKE to terminate until any replacement
    methodology has acquired credibility

MC - Principles for Measurement
  • (1) use income based measure of resources
  • (2) complement with small set of specific
  • (3) draw poverty lines with regard to social
  • (4) establish a regular monitoring system and
    update poverty lines within a five year period
  • (5) set a both a backstop and a target by using
    fixed and moving poverty lines
  • (4) (5) less important if real growth in median
    is small
  • (6) offer leadership and build public support for
    poverty reduction.
  • Explicit Link to Human Rights Agenda would be

LO1 Annual Income ?
  • Year compromise accounting period
  • immediate needs are much more short term
  • long term impacts may be multi year
  • Implicit Assumption income is perfectly
    substitutable within period but zero between
    period substitutability
  • Consumption Income
  • smooth consumption despite low income ?
  • Borrow?
  • Beg from relatives ??
  • BUT credit access for most low income households
    is very poor families may be poor /or alienated

Why does poverty duration matter?
  • long-term poor are materially worse off than
    short-term poor
  • Income-based measures imperfect indicator of
  • Low current income but buffers access to
    credit, run down consumer durables, obtain
  • long spells exhaust buffers to consumption
  • Psychological/Social impacts may increase with
  • individual duration
  • Social stereotypes may depend on the distribution
    of durations
  • High turnover/short spell process means more of
    nonpoor have been there

Why does immediate need matter ?
  • No groceries no credit no income immediate
    need in a week or less
  • Transitory child poverty can imply events
  • Events can have long run impacts
  • Eviction from housing ?
  • Forced mobility school attainment ?
  • Family stress gt Family breakup? Abuse ?
  • Traditional Welfare State agencies recognized
  • But new HK paradigm has firmly long run
  • Much higher incidence (3x in SIPP) than in annual
    poverty measures

(No Transcript)
LO2 Poverty Gap matters too !!Why not measure
Poverty Intensity ?
  • Sen-Shorrocks-Thon (SST) Index.
  • P (RATE ) (GAP ) (1G (X )).
  • Empirically inequality of poverty gaps 1G
    (X) is very nearly constant.
  • SST volume of Box, one dimension nearly
  • Changes over time, differences across
    jurisdictions can be approximated by.
  • RATE x GAP FGT index (a 1)
  • Poverty Box in 2 dimensions.
  • Poverty Intensity proportional to area RATE X

(No Transcript)
Variation over time mostly in rate, but misses
recent trend in poverty gap
Qualitative Conclusions Re Poverty Trends Often
Differ if consider gap
  • Relative poverty in early 1990s ?
  • Canada, UK, Sweden poverty rate gap disagree
  • US Germany indices agree
  • UK Absolute poverty line - poverty rate and
    poverty intensity change in opposite directions
    every time - 1974/79/86/91/95

(No Transcript)
(No Transcript)
(No Transcript)
(No Transcript)
(No Transcript)
Communicability ?
  • The whole point of measuring poverty is to
    influence the policies that might affect poverty
  • A poverty index not used or understood by the
    public by policy makers is pointless
  • Humans not very good at deriving accurate
    relative size information from large tables of
  • Quite good when presented as area proportionate

Equivalence scale - a matters