Here

advertisement
Working with household income variables
Practical computing session I – Part 3
Heike Wirth
GESIS – Leibniz Institut für Sozialwissenschaften
DwB-Training Cource on EU-SILC , February 13-15, 2013
Romanian Social Data Archive at the Departement of Sociology
University of Bucharest, Romania
• EU-Silc is the most important data source regarding
comparative description and analysis of poverty across the
European countries
• Headline indicator AROPE (at risk of poverty or social
exclusion) based on
1.
2.
3.
2
At risik of poverty
Severe material deprivation
Very low work intensity
• highest AROPE rates: BG, RO, LV
• lowest AROPE rates: CZ, SE, NL
3
At risk of poverty (HX080)
• less than 60% of the national median equivalised household
income
Severe material deprivation (HX110)
• couldn’t afford at least 4 out of 9 material deprivation items
(HS010 – HS110)
Very low work intensity (RX050)
• ratio of workable months and actually worked months < 0.2
 All three indicators are included in the data as well as the
combined indicator AROPE (PX070)
 In the following the focus is on ‘at risk of poverty’ indicator or
rather its components
4
Household income - Annual income from all sources
• Less taxes & social contributions
Disposable household income
• account for difference in household size (Equivalence measures)
• per capita income (each household member gets the weight 1)
• LIS-Scale (Square root of number of number of hh members )
• OECD Scale (adjusting for hh size & age of hh members):
• first adult (>= 14) = 1
• each additional adult ( >=14) = 0.5
• each child (< 14) = 0.3
Equivalised disposable household income
• Poverty line (Threshold of poverty)
• e.g. 40%; 50%, 60%,70% of the national median equivalised disposable income
At risk of poverty = income below the poverty line
5
Household income - Annual income from all sources (HY010)
• Less taxes & social contributions
Disposable household income (HY020)
• account for difference in household size (HX040) (Equivalence measures)
• per capita income (each household member gets the weight 1)
• LIS-Scale (Square root of number of number of hh members )
• OECD Scale (adjusting for hh size & age of hh members) (HX050):
• first adult (>= 14) = 1
• each additional adult ( >=14) = 0.5
• each child (< 14) = 0.3
Equivalised disposable household income (HX090)
• Poverty line (Threshold of poverty)
• e.g. 40%; 50%, 60%,70% of the national median equivalised disposable income
At risk of poverty = income below the poverty line (HX080)
6
Some Questions regarding Poverty
• Does poverty vary widely between different country in Europe
• How is income poverty correlated with gender, age, education,
urban and other characteristics?
• What are the main sources of income for the poor?
• How does unemployment affect poverty?
• How is income poverty linked to health conditions & access to
health services?
• Consistency of poverty
How long people are poor?
• How often people are poor over a period?
•
7
Types of exploring poverty profiles
• By poverty status: Poor versus non-poor
•
Summarize the distributions of characteristics, e.g. household
type, household size, activity status, educational level,
urbanisation degree of residence, age …
• By characteristics: e.g. age, health condition, number of
children, educational level, occupation of household
head, region of residence.
•
8
Report the poverty rate for each components.
Hands on – Good to know
• All income variables in the UDB are in € (EURO)
• For countries not members of the euro area the conversion factor
can be found in variables HX010 & PX010.
•
•
•
•
Income data (euro) i. e. HY020 * HX010 = income data (national
currency)
HY010 – Total household gross income
HY020 – Total disposable household income
HY022 – Total disposable HH income before social transfer
These variables do not include PY080 although published indicators
were calculated with PY080
• Remember PY080G was summarized across household members and
added as the variable 'pens' to the household data
•
9
Please note:
• HY025 =Within-household non-response inflation factor
• Factor by which it is necessary to multiply
the total gross income (HY010);
• the total disposable income (HY020) or
• the total disposable income before social transfers (HY022)
•
to compensate the non-response in individual questionnaires
 If the data are not corrected by the effect of non-responding
individuals within a household, income of individuals not
interviewed is not added up into the total household Income.
 For further details see guidelines
10
Hands on – (for more hints see the exercise sheet)
***********************************************************.
* Exercise 3.1a
*
* Compute the equivalised household income (OECD scale) by yourself,
* name the new computed variable: HX090_new
*
*
* Check your results by comparing the means of HY020, HX090_new, HX090
* by HH type(HX060), &
* by country (HB020)
*
* Did you get what you expected?
***********************************************************.
11
Hands on – Exercise 3.1a
* open the household dataset.
GET FILE='data_path/household_data.sav'.
compute HX090_new = ((HY020+PENS)* HY025)/HX050.
var lab HX090_new 'oecd scale'.
mean tables HY020 HX090 HX090_new by HX060/cells = mean .
mean tables HY020 HX090 HX090_new by HB020/cells = mean .
12
Hands on - Exercise 3.1b
************************************************************************************.
* Exercise 3.1b
*
* Compare the means of the equivalised Income according to the OECD Scale
* with equivalised Income based on
*
* Per-Capita scale (divide disposable income by number of household members), and
* LIS-equivalence scale (divide disposable income by square root of number of household members)
*
* across Household Type
* across countries
* what are the main findings?
*
* Compute the mean of the equivalised disposable income depending on the
* number of unemployed persons (unemp) in the household, by country
*
* Collapse categories for number of unemployed persons: 0, 1, >=2
***********************************************************************************.
13
Hands on - Exercise 3.1b
* Per capita income.
compute pc_income = ((HY020+PENS)* HY025)/HX040.
* SQUARE ROOT EQUIVALISED HOUSEHOLD-INCOME.
compute HX040_2 = sqrt(HX040).
compute lis_income = ((HY020+PENS)*HY025)/HX040_2.
var lab pc_income 'per capita income'/
lis_income 'squareroot hhsize'.
means tables HX090 pc_income lis_income by HX060 HB020/ cells = mean .
temporary.
select if (HB020 ne 'DK').
means tables HX090 pc_income lis_income by HX060 HB020/ cells = mean .
* Equivalised Income by number of unemployed persons in the household *.
recode unemp (0=0) (1=1) (2 thru highest = 2)into unemp_n.
val lab unemp_n 2 '>=2'.
* check your new variable *.
cross unemp by unemp_n.
means tables = hx090 by unemp_n by hb020/cells = mean.
14
Hands on - Exercise 3.1c - Optional
***************************************************************************
*
* Now compute the equivalised income based on the disposable income before social
* transfers excl. old-age-benefits (HY022)
* Name the new variable HY22_equi
*
* Remember to take into account the variable 'pens' as well as the inflation factor (HY025).
*
* Compare the means of HX090 and HY022_equi by country
* What are the main findings?
*
* Compute the mean of the equivalised disposable income (before social transfers) by
* number of unemployed persons (unemp_n) in the household, by country
**************************************************************************.
15
Hands on - Exercise 3.1c - Optional
* equivalised income before social transfers *.
compute HY022_equi = ((HY022+PENS)* HY025)/HX050.
means tables = HY022_equi HX090 by HB020/ cells = mean .
means tables = HX090 HY022_equi by unemp_n by hb020
/cells = mean.
16
Exercise 3.1d - Optional - At Risk of Poverty
************************************************************************************
* The at-risk-of-poverty rate is the share of people with an equivalised disposable income
* (after social transfer) below 60 % of the national median equivalised disposable income after social transfers
* This poverty indicator is included in the EU-SILC data (HX080).
* Below you find an example of how to compute this poverty line
*
Step 1: compute the national median, name it median_sample *.
aggregate outfile = * mode = addvariables overwrite = yes
/break = HB010 HB020
/median_sample = median(HX090).
means tables median_sample by hb020.
step 2: compute the poverty line, name it povline_60 *.
compute povline_60 = median_sample * 0.6.
step 3: compute the poverty- indicator - name it at_risk_60 *.
compute at_risk_60 = 0.
if (hx090 lt povline_60) at_risk_60 = 1.
17
Exercise 3.1d - Optional - At Risk of Poverty
****************************************************
* compare the findings with poverty indicator (HX080) provided in
the data
• Reason for the discrepancies ?
• *
* Compute the percentage of households at risk of poverty
depending on the number of unemployed persons (unemp_n) by
country
*
***************************************************.
* % of households at risk by number of unemployed persons in hh, by
country *.
cross unemp_n by HX080 povline_60 by HB020 / cells = row.
18
Exercise 3.1e - Optional - Make ends meet
************************************************************************************.
* Subjective indicators regarding the financial situation of the household
* HS120: Ability to make ends meet (scale 1-6: with great difficulties (1) ... very easily(6))
* HS130: Lowest monthly income to make ends meet
*
* Compute the average ability to make ends meet depending on whether a household is at
* risk of poverty or not, by country
*
* Focus on households with difficulties or great difficulties to make ends meet.
* Compare the average of the lowest monthly income needed (HS130) to the disposable income, by
* country.
*
* Do the same for households which make ends meet easily or very easily
* Main points ?
************************************************************************************.
19
Exercise 3.1e - Optional - Make ends meet
* average ability to make ends meet by risk at poverty & country *.
means table HS120 by HX080 by HB020 / cells = mean.
* Which income variable seems approriate for this comparison?.
* Equivalised income (HX090)
* Total disposable income (HY020) respectively the published HY020
* HY022_pens = ((HY020+PENS)* HY025)
* what else ?.
compute monthly = (HX090/12).
compute HY020_pub = ((HY020+Pens)*HY025)/12.
* compute HY020_m = (HY020/12).
* All households *.
means HX090 monthly HY020_pub HS130 by HB020 /cells = mean.
20
Exercise 3.1e - Optional - Make ends meet
* only households with difficulties or great difficulties *.
compute difficult = (HS120 le 2).
compute easily = (HS120 ge 5).
val lab difficult 1 'difficulties' 0 'no difficulties'.
val lab easily 1 'easily'
0 'not so easily'.
cross hs120 by difficult easily.
filter by difficult.
means monthly HY020_pub HS130 by HB020 /cells = mean.
filter by easily.
means monthly HY020_pub HS130 by HB020 /cells = mean.
filter off.
21
Exercise optional
*******************************************************.
 Calculate Income Quintile for households, by country
 Calculate the mean of equivalised disposable income for the
lowest & highest income quintile, by country
 Calculate the mean of lowest monthly income to make ends meet
by income quintile & country
 Calculate percentage of households with difficulties/great
difficulites to make ends meet, by income quintile and country
 e) Severe material deprivation rate, by income quintile and
country
*******************************************************.
22
Exercise optional
*compute the quintiles for each country based on HX090 *.
* Option 1 - use the comand split file *.
split file by hb020.
RANK VARIABLES= HX090
/NTILES(5) INTO Quintile
/print = yes.
freq quintile.
split file off.
23
Exercise optional
************************************************************************
Option 2 - Rank variables (however HB020 = string variable)
=> use HB020_num
************************************************************************.
autorecode variables = HB020 / into HB020_num.
RANK VARIABLES= HX090 by HB020_num
/NTILES(5) INTO Quintil.
cross HB020 by quintil.
* mean of equivalised disposable income for the lowest & highest income quintile, by country *.
Temporary.
select if (quintil = 1 or quintil = 5).
means tables hx090 by hb020 by quintil
/cells = means.
24
Exercise optional
* Calculate the mean of lowest monthly income to make ends
meet by income quintile & country *.
means tables HY020_pub HS130 by HB020 by quintile
/cells = means.
* difficulties/great difficulites to make ends meet, by income
quintile and country *.
cross hb020 by difficult easily by quintil / cells = row.
25
You did it …
… at least
for today
Source:
: http://www.hierschreibenwir.de/node/32556
26
Types of equivalence measure
>= 14
>=14
<= 14
Household size: 4; income 4000 €
Per-capita income = 4000/4 = 500 €
Lis-equival. income = 4000/Square root(4) = 2000 €
OECD (new)=4000/(1+1*0.5+2*0.3)=4000/2.1= 1905 €
Source:
27
: http://www.hierschreibenwir.de/node/32556
If not taking HY25 into account
2000 €
2000 €
0€
Household size: 4; income 2000 €
Per-capita income = 2000/4 = 1000 €
Lis-equival. income = 2000/Square root(4) = 2000 €
OECD (new)=2000/(1+(1*0.5)+(2*0.3))=2000/2.1= 1905 €
Source:
28
: http://www.hierschreibenwir.de/node/32556
Download