Measurement session 5

1
Measurement session 5

• Wages and Income

2
Issues in mst of wages and income

• Wages individual
• Self-employment, capital how to measure their
  income?
• Other (non-monetary) resources?
• How to compare households of different size and
  composition?
• Income a household level variable?

3
Issues in mst of wages and income

• Why do we want to measure income?
• We assume income money a universal
  equivalent with which each consumer can buy
  whatever s/he wants
• Should we not be measuring consumption, or
  capabilities? Or (subjective) welfare/happiness?

4
Income or consumption?

• Consumption reflects income as well as past
  savings, access to credit markets, and seasonal
  variation in income
• No records of income or seasonal fluctuations
• Large informal sectors
• Consumption data helps in deriving the poverty
  line

5
Income or consumption?

• Measuring income is a theoretical, ethical
  choice income is supposed to offer potentially
  anything, whereas consumption is the outcome of a
  choice
• Ita a liberal point of view we want to measure
  situations before choices, and not care about the
  outcomes

6
The measurement time-frame

• There is no good reason to choose the year as
  unit
• Some consensus that poverty is real when
  deprivation for 3 years (consumers smooth their
  consumption, save or borrow)
• But panel data is so noisy it doesnt improve
  measurement!

7
Data needs for poverty analysis

• National level data
• National accounts GDP, consumption, savings,
  investment, imports, exports, etc.
• Ministry of Finance, Central Statistical Agency
• Budgets, price surveys, and data collection
• Monthly, quarterly, and yearly

8
Data needs for poverty analysis

• Household Individual level data
• Household income, consumption, employment,
  assets, production, demography, etc.
• NSIs, sectoral ministries, NGOs, academics
• Household survey, rapid assessments, monitoring
  and evaluation
• Yearly, 2-3 years, every 5 years

9
Available data

• Administrative data taxes and payrolls, mainly
• Population Census
• Household surveys Labor Force Survey, HBS, SILC
  (European Panel)
• Qualitative and Participatory Assessments
  ethnographic, village studies, beneficiary
  assessments, etc.

10
(1) Wages

• Whats in a wage?
• Much is excluded all the in-kind payments
• Time frame hourly, monthly, yearly wage?
• Statistical sources
• Wages employers tax declarations
• Wages and employment Labor Force Survey
• Income households income tax declarations

11
(1) Wages average annual wages (full-time
workers)
Private sector Civil servants
12
(1) Wages annual earnings

• But define earnings as the sum of wages earned
  over one year by all those who have worked at
  least 1 day
• The diagnosis is quite different

13
(1) Wages annual earnings

• That is because earnings are a composition of
  wages earned number of days worked
• Here are the average number of worked days for
  men and women

14
(1) Wages annual earnings

• Earnings are a composition of wages earned
  number of days worked
• Here are the average number of worked days for
  men and women, by age group

15
(1) Wages annual earnings

• The average yearly wage from national accoutning
  sources can also look very different depending on
  the numerator and denominator you choose

16
(1) Wages annual earnings

• National accounting
• all wages/ average labor force (number of
  employed at beginning of year at end of year) /2
  
• Alternative denominator all those who have
  worked at least 1 day during the year
• Gross wages have increased much more than net
  wages because taxes on wages have increased
• The diagnosis is definitely not the same!

17
(2) Other income

• Self-employed
• Taxes are an unreliable source
• Depend on the legal status of the business
• More fundamental problem for themselves, there
  is no conceptual difference between their
  household budget and their business

18
(2) Other income

• Survey data
• Finally, European comparisons are made from
  survey data
• SILC survey on income and living conditions
• Many questions on income
• Survey effect the more numerous the questions,
  the richer the respondents!

19
Whats in an income

• Things that are ill-measured
• Income from capital (wealth) would have little
  impact on poverty since almost entirely above the
  median
• Yet changes a lot when considering inequality

20
Taxes

• Again, not simple
• Income tax is removed from disposable income
• But what about local taxes?
• Again, it depends on what you consider a choice
  or not
• Ex is living in Paris a choice? Yes ? you use
  your income to pay local taxes. No ? local taxes
  should be removed from income

21
Transfers between HH

• Alimonies and the money transfers of migrants to
  their homeland are removed from disposable income
• But it may underestimate income migrants send
  money for their own future use, too

22
Whats in an income

• Choices of what to include are often made for no
  good theoretical reason but practical ones

23
Recent pushes towards better income measurement

• French official report (CNIS) general Eurostat
  tendency
• Goal take better account of non monetary
  resources

24
Recent pushes towards better income measurement

• Imputed rent
• Owning your home

25
Recent pushes towards better income measurement

• Home production

26
Recent pushes towards better income measurement

• Public goods (that can be individualized)

27
Niveaux de vie

• Dabord faut le définir
• Income (gross net including non market
  goods?)
• By consumption unit

28
(1) Wages annual earnings
29
Equivalence scales

• Are used in setting level of allowances
• Ex  RMI  in France
• 425,40 for 1 single person
• 638,10 for 1 couple
• 765,72 for 1 couple 1 child.
• Underlying hypothesis 425,40 buys same quality
  of life when single than 765,72 when 2 parents
  1 child
• 638,1 425,40 0,5 425,40
• 765,72 425,40 0,5 425,40 0,3 425,40
• Implicit equivalence scale 1st adult 1 2nd
  adult 0,5 child 0,3

30
Equivalence scales Where do they come from?

• No consensus example of number of consumption
  unit / child?
• Household budget surveys direct child expenses
  8. But how to split food, housing expenses?
  -gt comparing parents / non-parents (but
  unobserved taste differences). Much noise!
• Lechêne, 1993 between .2 and .7
• Subjective measurements (how much do you need
  to) -gt even wider dispersion

31
Equivalence scales example on 4 variants

• The 4 variants

Source Du bon usage des échelles déquivalence
Limpact du choix de la mesure, Jérôme Accardo.
http//www.cairn.info/revue-informations-sociales-
2007-1-page-36.htm
32
Equivalence scales example on 4 variants

• Results

All figures calculated on fiscal data for
2001 Source Du bon usage des échelles
déquivalence Limpact du choix de la mesure,
Jérôme Accardo. http//www.cairn.info/revue-inform
ations-sociales-2007-1-page-36.htm
33
Equivalence scales example on 4 variants

• Composition of the poor depending of
  equivalence scale used
• From top to bottom
• Couple w. children
• Couple w.out children
• Single parent w. children
• Single adult

Source Du bon usage des échelles déquivalence
Limpact du choix de la mesure, Jérôme Accardo.
http//www.cairn.info/revue-informations-sociales-
2007-1-page-36.htm
34
Intra-household allocation

• Duflo and Udry (2004) NBER Working Paper 10498
• Data the Côte dIvoire Living Standards
  Measurement Survey (CILSS). 1985-1988. 1,500 HH
• Some crops are cultivated by men, others by women
• They do not benefit equally from rain

35
Intra-household allocation

• Results
• Rainfall shocks associated with high yields of
  womens crops shift expenditure towards food
• Rainfall-induced fluctuations in income from yams
  are transmitted to expenditures on education and
  food, not to expenditures on private goods
• Other crops fluctuations are associated with more
  consumption of private goods

36
Intra-household allocation

• Evidence from sociology in the US and France
• Income is not 100 shared among household members
• There are intra-household variations in
  disposable income and consumption
• But we know way too little to take them into
  account statistically so far.

37
In 2001 in a few European countries
38
Lessons to be learnt

• On child poverty policy impact of measurement
  linked with ideology of those you want to
  convince (figures for advocacy)