Title: Ec355, lecture 5 actually 4'5: Poverty and Social Policy Lecturer: David Reinstein
1Ec355, lecture 5 (actually 4.5) Poverty and
Social Policy Lecturer David Reinstein
2This week
- Poverty and Social Policy
- Defining and measuring poverty
- Poverty in the UK (and other wealthy countries)
- Social Policy
- Policies to combat poverty (in general)
- The UK welfare system
- Global Poverty (Besley and Burgess paper)
- Measurement, patterns, trends
- The role of policy
- Cleaning up
- Term paper
- Exercises
- Pizza tax additional intuition
- Discussion
- Student evaluation of teaching
- Reading CM chapter 15 (skim 15.4 and 15.5),
17.1-17.2, skim 17.3, read 17.5. Besley and
Burgess (2003)
3Defining poverty
- When needs exceed resources
- Needs for survival? For optimal health?
- Attainment of minimum capabilities (health,
transportation, social acceptance, education) - Needs for full participation in society?
- Should our measure depend on individual choices?
- A relative measure?
- Economics some minimum level of utility?
- Again, the relevant measure depends on its
intended use! The question which measure is
better cannot be answered in a vaccum. - discuss
4Poverty lines
- A poverty line identifies a level of income
considered to be necessary to prevent poverty. - Difficulty Different costs, requirements under
different conditions and societies, public
services matter. - The poverty line may have to be adjusted for
local purchasing power and background amenities.
5Absolute poverty line
- Rowntree (1901) The level of income necessary
for the bare necessities of life. - measures poverty in terms of lack of command
over basic consumption items that are essential
for maintaining physical efficiency now called
the cost of basic needs (CBN) method of setting
poverty lines (Ravallion and Bidani 1994). The
poverty line is the sum of food and nonfood costs
in a basic consumption basket. (Kakwani, 2003)
6 determined by physical needs
7Relative poverty
- Level of income that enables participation in
society. - A subjective measure. Townsends index gained
influence in the 1980s. Criticised for being
based on individual choices.
8 - 2008 A single person in Britain needs to earn at
least 13,400 a year before tax for a minimum
standard of living, the Joseph Rowntree
Foundation says. (see singleadult_minimum.pdf
for more details (on CMR))
Note that a commonly used measure of poverty is
having income less than half of the national
average (i.e., below about 12,000/year per
capita in the UK). Compare Global poverty line
defined given as roughly 1.25 per day (about 6,
300 per year). The second tier benchmark of
about twice this (derived from the average of
national poverty lines in lower middle income
countries) covers a third to half of the world.
9Outcome measures of poverty
- Sen (1983) 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
characteristics - For richer communities, however, the
nutritional and other physical requirements
were typically already met, and the needs of
communal participation would have a much higher
demand in the space of commodities - Possible critique Couldnt the person in the
richer community join a poorer one instead? - (Townsend Even the core capabilities are
relative to a particular society)
10Measuring poverty as P(y,p) Sens (1976) Axioms
- Focus changing the income of someone above the
poverty line should not effect P - Monotonicity axiom P should rise (other things
held equal) if the income of a poor household
falls (some element of vector y declines). - Weak transfer axiom A transfer of income from a
poor household to a poorer household should
decrease P (and vice/versa). - Anonymity/symmetry labeling doesnt matter
- ?what about the number or ratio in poverty?
11Measures
- Poverty headcount ratio fraction of population
below poverty threshold. Fails monotonicity and
weak transfer axioms. - Hq/n
- Poverty gap (average) distance below the
poverty line for those in poverty. Fails weak
transfer axiom.
12Sens index (1976)
- Where q is the number in poverty, n the total
population, G is the Gini coefficient among the
poor, z is the poverty line, yi is (poor person)
is income, hence gi her income gap, and I the
average income gap (S(z) is the set of poor
people). -
- Sens P satisfies the previous axioms plus some
more that Sen thought important. - As Sen shows, this measure in effect, weights
income gaps by rank in some interpersonal welfare
ordering. Note if the poor are all equally poor
(G1) then only the income gap HI matters. - (I see no clear connection between this index and
capabilities.)
13Poverty in the UK (and other wealthy countries)
14(No Transcript)
15(No Transcript)
16(No Transcript)
17Social Policy
- Policies to combat poverty (in general)
- Public goods and services with progressive
taxation pro-poor - Regulation and mandates (e.g., minimum wage
legislation, affirmative action
anti-discrimination programmes) - Benefits (and income redistribution) ...
- Other
- Encouraging charitable giving and volunteering
(tax benefits) - Government employment, Birth control policies
(controversial), (Developing countries)
provision of micro credit, improved property
rights. - direct and indirect
18Benefit issues
- Cash payments versus targeting neccesities (food
stamps, council housing) and merit goods
(education, medicines). Class discussion of pros
and cons - Argument for cash 1. With rational actors who
also are acting in their (own and childrens)
best interest a lump-sum transfer is optimal (see
2nd welfare theorem). 2. government provision of
goods or directed benefits are vulnerable to
other inefficiencies, distortions, and fraud. - Argument for other 1. Lump-sum transfers are
infeasible perhaps in-kind transfers have less
distortionary effect then (e.g., a means-tested
benefit). 2. Poor people may need to be guided to
merit goods perhaps they are more vulnerable
to the pathologies discussed in behavioral
economics and psychology such as myopic
preferences. 3. Benefits are given to parents
they may not act in their childrens best
interests. 4. Externalities.
19- Contingent benefits target an identifiable group
(e.g., the retired, single mothers, or low-caste
Indians) more likely to experience poverty versus
income-related (means-tested) benefits. - Argument for former Less tied to behaviour, so
less distortion of the labour supply decision.
May be cheaper to administer. - Argument for latter With a positive mcpf
(remember that?) the less spent the better.
Means-testing can better target those in real
poverty.
20Digression to labour market (ch 13)
- Labour-leisure tradeoff (depict indifference
curves) - Budget constraint (leisure versus consumption)
determined by wage rate (net of taxes) slope in
hours worked (depict) - Impact of proportional income tax A decrease in
the (net) wage rate will lead to an decrease in
the labour supply only if the substitution effect
dominates the income effect.
21(No Transcript)
22(Dis)incentives Poverty and unemployment traps
- (refer to fig. 13.4 and box 13.4)
- Unemployment benefits (and wealth and informal
means) lead to a minimum level of income (or
consumption) even with no hours worked. This
point of discontinuity means some will choose not
to work. Depict adjusted budget constraint. - The rate of benefit withdrawal can mean that
purchasing power declines as hours worked
increases over some range here no one would
want to work (a number of hours in this range).
This is a poverty trap. - An unemployment trap is where this occurs for the
first hour worked. -
- Depict both
-
23(No Transcript)
24The negative income tax (will fix)
- Designed to give more incentive to work, remove
poverty traps (counter benefit withdrawal) - CM depicts a simple lump-sum benefit with
subsequent proportional taxation this is
essentially an administrative change only a
real negative income tax is different. (I just
realised this, will give a corrected picture for
next week) - Subtle problem to have progressive taxation, at
some point rate must increase, and negative
income tax can make this jump more abrupt. - (details) and graphs here
- Ynet Yt-t(Yi-YA)
25Minimum wage legislation
- The standard argument is that it under perfect
competition it creates unemployment and
inefficiency, merely shutting the low-skilled out
of the market. - But with monopsony power (e.g., a single employer
and many independent workers) a price floor can
improve efficiency. depict on board At a
higher wage more people are willing to work. And
even if employment is unchanged this should lead
to lower inequality.
26(monopsony power)
- MRP is the Marginal Revenue Product.
- S is the elastic supply of labour.
- MFC is marginal factor cost this includes both
the wage paid and the increment to previous wages
needed to hire another unit of labour. - Without a minimum wage the monopsony firm sets
MFCMRP at level Lm yielding wage 4 (on the
supply curve. - With the minimum wage the monopsony firm sets
MRPminimum wage 5 and hires L2 in labour.
The deadweight loss of monopsony is also reduced.
27- Empirical results are mixed. Discuss Card, etc.
- Fuchs et al. (1998) polled labor economists at
the top 40 research universities in the United
States on a variety of questions in the summer of
1996. Their 65 respondents split exactly 50-50
when asked if the minimum wage should be
increased. - Minimum wage is a blunt tool may help in some
sectors, hurt in others, especially those that
can switch to cheaper labour markets.
28(No Transcript)
29Minimum wage introduced in UK in 1999
- Currently 5.73/ hour (compares, e.g., to 4.31
in USA) - The Low Pay Commission found that, rather than
make employees redundant, employers have reduced
their rate of hiring, reduced staff hours,
increased prices, and have found ways to cause
current workers to be more productive (especially
service companies).
30The UK welfare system
- Major expansion of the welfare state in the
1940s, some scaling back in the 1980s. - Job seekers allowance contribution based and
income-based (basic rate) (see JobCentre for some
details note terminology has changed since CM)
. Must show you are actively seeking work. - Income support (alternate) means-tested for
groups not required to seek work (pensioners,
single parents, disabled) - Retirement pension
- Housing and council tax benefit, social fund
- Child benefit
- Working tax credit
- Incapacity and disability benefit
- Welfare to work (education and training scheme
direct government provision/subsidy)
31Government (welfare) spending
32UK benefit rates
- See http//www.jobcentreplus.gov.uk/JCP/stellent/g
roups/jcp/documents/websitecontent/dev_015666.pdf)
- For single person over age 25
- JSA max 64.50
- Income support same
33(No Transcript)
34Replacement rates (updated to 2003, OECD)
35Global Poverty (Besley and Burgess paper)
- Although the fraction in poverty is falling, the
absolute numvers have shown limited change - Millenium Summit (2000) ambitious set of goals,
including reducing the proportion of people
living on 1 a day from 30 to 15 of the worlds
population by 2015. - Our agenda for reducing poverty needs to be
built on firm theoretical foundations why? - How to achieve this?
36Poverty line
- Chen and Ravillion (2001) use price and
consumption basket data fom the 1993
International Comparison Project to come up with
the 1.08/day figure Also reflects domestic
poverty lines. - A conservative measure for middle income
developing countries.
37Description Who the poor are and how they live
- Great heterogeneity by region, country, and
within countries. - The Chinese Miracle and the African Tragedy
(next slide) - Cross-country analysis. Difficulties in
comparability. Signpost for future work/
combining macro and micro findings.
38(No Transcript)
39(No Transcript)
40Variations in inequality
- Changes in income/capita do not seem correlated
to changes in inequality (at least in this
context) stability within countries over time. - Regression relation to poverty differences
41(No Transcript)
42Role of policy
- Previous Washington consensus sound fiscal and
monetary policy, privatization, property rights,
free trade. - Change in economic thinking/focus
- Institutional and political context crucial
consider elite and political capture - Need for pro-poor policy, particularly in
credit and property rights - Land reform (tricky)
- Need to focus on distributional impact of growth
- Redistribution though taxes is difficult
- Social infrastructure important
43Credit
- Pro-poor expanded credit aggregate credit
expansion not enough. - Argument depends on particular asymmetric
information/moral hazard story. - Microcredit peer selection and monitoring,
typically operated by NGOs. Why do they have
such high repayment rates.
44Property rights (and land reform)
- Equity and efficiency?
- Means to access credit
- Alleviating moral hazard
- But unintended consequences
- Reputation and networks as substitutes for formal
enforcement - Duflo The poor are not given capital optimally
because they cannot make binding comittments
45Last weeks questions
- CM chapter 14 (Q 2 3)
- Q. 2 What advantages/disadvantages does the Gini
coefficient have over other measures of equality? - Advantages A single number (unlike Lorenz
curves), widely used, interpretable graphically
(via Lorenz curves). Satisfies the anonymity and
relativity axioms. - Disadvantages Hard to interpret as a welfare
measure the weights placed on each individual
are fairly arbitrary. May underemphasize the
income distribution at the bottom end, which is
more relevant to extreme poverty. Is not as
elegantly decomposable into within-group and
between group inequality (as in Theils entropy
index).
46- Q 3 An index of inequality is constructed based
on unequivalised individual incomes at a
particular moment. Would the index be higher or
lower if (a) lifetime income was used (b) if
equivalised income was employed and (c) if
household income was used instead? - a. Lower, as income fluctuates throughout the
life cycle, adding variance to the measure. - b. So, the texts discussion of equivalence
scales is confusing. In the literature not
using equivalence scales means assuming that
everyone in a large household (with total income
Y) gets the same income as a singleton with the
same income Y. So it is as if everyone in the
household had income Y. This means basically
assuming that all goods in the household are
non-rival (like heat). Using equivalised income
means adjusting the incomes downward for people
in larger households. As larger households tend
to be wealthier, adjusting these incomes
downwards (relative to smaller households) will
lead to a lower measure of inequality, at least
at first. But if we keep adjusting it downwards
(towards the extreme in which each individual in
a household effectively gets only a 1/N share of
the income) the equivalized income of the
(individuals in the) larger households eventually
goes below that for lower households, and our
measure of inequality increases again. This
argument depends on per-capita income (the 1/N
measure) being lower in larger households. - c. This question is hard to interpret, sorry.
47New homework
- CM exercises 15.2, 15.4 (based on recent recent
statistics), 17.1 (explain in context of poverty
trap), 17.3 (give arguments in both directions.
Also 13.5. - Read CM ch 4-5, also (added) 13.1-13.2 (ignore
Slutsky, ignore reservation wage, dont worry
about real income versus money-income effect.
Know the graphical analysis), 13.4 benefit traps.