1
InGRID Winter School in Cross-country
Microsimulation
Day 1
Silvia Avram, Chrysa Leventi, Iva Tasseva, Alberto
Tumino
ISER, U Essex
18-20 February 2015
Course based on EUROMOD v. G2.0+
2
Outline: Day 1
• Morning (9:30-12:30 with a break at 11:00)
– Tax Benefit Microsimulation and EUROMOD
– Model Design
– The EUROMOD User Interface(UI)
– Running EUROMOD
– Summary Statistics Tool
– Error handling Documentation
• Afternoon (14:00-17:00 with a break at 15:30)
– Hands-on practice: Exercise 1
– EUROMOD functions and parameters
– Hands-on practice: Exercise 2
3
Outline: Day 2
• Morning (9:30-12:30 with a break at 11:00)
– EUROMOD Functions and Parameters (cont.)
– Hands-on practice: Exercise 3
– EUROMOD Functions and Parameters (cont.)
– Income Lists
• Afternoon (14:00-17:00 with a break at 15:30)
– Hands-on practice: Exercises 4 & 5
– Hands-on practice: Exercise 6
4
Outline: Day 3
• Morning (9:30-12:30 with a break at 11:00)
– Assessment (tax) units in EUROMOD
– Hands-on practice: Exercise 7
• Afternoon (14:00-17:00 with a break at 15:30)
– Hands-on practice: Exercises 8 & 9
– Presentations by participants
– Using EUROMOD after the course
– Q&A
5
Tax-benefit microsimulation and
EUROMOD
6
Microsimulation
• Micro: technique using units (ex: individuals, households, firms etc.)
instead of aggregate information
• Simulation: application of an intervention that may change the state or
behaviour of units
• Contribution: estimates results derived from the application of these rules
on each unit.
7
Tax-benefit models
• Micro: household micro-data
• Simulation: taxes and benefits
• Contribution: impact on disposable income
.75
Why tax-benefit microsimulation?
0
.25
.5
• Population diversity
and frequency
• Policy complexity:
detail and interactions
1
2
3
4
5
deciles
6
7
8
9
10
Source: Euromod using EU-SILC 2004
9
What makes EUROMOD special?
• Multi-country tax-benefit model for the European Union: unique
• Harmonised data and simulations
• Very flexible structure (but scope depends on data available)
• Tax-benefit modelling language: universal
• Library of policies
10
What can EUROMOD do?
• Simulate previous, current, future and “potential” tax-benefit rules
– Distributive analysis
– Budgetary effects
– Indicators of work incentives
• Complex policy reforms (e.g. revenue-neutral)
• Policy swapping
• Counterfactual (“what if”) scenarios (e.g. stress test)
• EU-wide policy reforms
• Tax evasion and non-take-up simulation/calibration (special data )
• Input in other analyses (ex: labour supply models; links with CGE etc.)
11
Examples-Counterfactual scenarios (I)
• To what extent tax-benefit systems support those who became
unemployed at the onset of the Great Recession?
Source: Fernandez et al. (2013) Review of Income and Wealth
12
Examples-Counterfactual scenarios (II)
• Poverty effects of child related policy instruments in Poland
35%
Baseline 2005
30%
25%
PL-2007
20%
Austrian system
15%
French system
10%
UK system
5%
No fam ben
0%
FGT0
FGT1
FGT2
Source: Levy, Morawski and Myck, Euromod Working Paper 3/08
13
Examples-Work incentives
• Calculate Marginal Effective Tax Rates
Source: Jara and Tumino (2013) International Journal of Microsimulation
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
National
EMU-UI flat
EMU-UI %
14
Examples-EU-wide reforms
• Net replacement rates with and without a EU-wide UI benefit
110
100
90
80
70
60
50
40
30
20
DE
EE
EL
ES
FR
p25-p75
IT
LV
Median
Mean
AT
PT
FI
Source: Jara & Sutherland (2014) Social Situation Monitor Research Note 3/13
.4
.2
0
EE05
UK01
PL05
SW01
FR01
FI98
PT01
US05
HU05
DK98
BE01
GE01
IE01
NL01
IT98
AT98
SP01
GR98
UK01
PT01
PL05
IT98
NL01
FR01
SW01
AT98
SP01
US05
HU05
EE05
FI98
GR98
BE01
GE01
DK98
IE01
.6
.4
.2
0
Own-wage elasticity
.8
Point estimate
PT01
HU05
PL05
EE05
FR01
DK98
AT98
NL01
GE01
SP01
FI98
SW01
US05
UK01
IE01
GR98
BE01
IT98
.6
Own-wage elasticity
.8
PT01
NL01
PL05
AT98
FR01
HU05
EE05
GR98
SW01
US05
GE01
UK01
IT98
DK98
BE01
FI98
SP01
IE01
15
Examples-Behavioural reactions
• Calculating labour supply elasticities
Married women
Married men
.8
.6
.4
.2
0
Single men
Single women
.8
.6
.4
.2
0
95% confidence interval
Source: Bargain et al. (2013) Journal of Human Resources
16
EUROMOD in this course
• EUROMOD is continuously being developed and improved
• Latest public release (G2.0+)
• Countries and policies
– BE, GR, IT, ES, UK, LT, CZ, HU, EE: 2005-2013
– NL, SE, IE, PT, CY, PL, SI, LV, SK, FR: 2006-2013
– DK, LU, FI, AT, DE, MT, RO, BG : 2007-2013
• Data
– EU-SILC data acess issue
– training data
17
Overview of EUROMOD design
18
Structure
Input microdata
(text file)
Simulations
(EURMOD
Engine in C++)
Policy parameters
(XML filesEUROMOD UI)
Output microdata with
additional simulated
variables (text file)
EUROMOD input dataset
• Variables: demographic, labour, income, assets, expenditure
• Harmonised data reference period
• Compulsory variables (e.g., id, age, weight, incomes)
• No missing values
• Gross income
• Monetary variables reported on (average) monthly basis
• Documentation (do-files template and DRD)
• Currently-based on SILC
19
Variable name convention
Names are combination of acronyms: abb**
a – type of information (e.g., y: income, x: expenditure)
bb – specific for each type a (e.g., y| em: employment, se: self employment)
eg.
yem: employment income
yse: self-employment income
** further bb’s for additional information/detail
eg.
ysebs: business self-employment income
exception id*, eg. idperson, idmother
20
21
EUROMOD policy parameters
• Contain all info about tax-benefit rules
• Stored in XML files read by the EUROMOD engine
• Two files per country
– Data config file
– Parameters file
• Common Variables file (VarConfig.xml)
• Manipulated via user interface (UI)
• UI-stand alone software based on .NET framework
• Implemented via EUROMOD functions grouped in policies
– General settings
– Defining elements to be used later on (tax units, income lists,
constants etc.)
– Simulation of policies
– Controlling the output file
22
EUROMOD
User Interface
23
Installation
• Requires Microsoft .NET framework files or an Internet connection to
download files in the SETUP process
• Complete separation between UI and ‘content’ (i.e. XML) files
• Only one copy of the UI but can use multiple ‘content’ files
• ...but content files must have set structure of folders
• Run the Installation Wizard
• Set the path to your EUROMOD files
– Project path
– (if necessary) separate input data and output data paths
24
EUROMOD folder structure
25
Linking EUROMOD to content files
26
User Interface (UI)
• Single stand-alone piece of software-Windows OS
• Single working environment
• Mostly point and click but some hot keys are available (standard and
specific)
• In-built features that allow for improved user control and guidance
• Intuitive!!
• Features:
– Ribbon bar with tabs
– Context menus
– IntelliSense (suggestion of parameter values )
– Drag and drop
– Bookmarks and comments
– Built-in help
27
User Interface (UI)
Ribbon bar
Run button
Country files
28
Working environment
open country
policy systems
comments
policies
29
Ribbon bar
Country tools:
-country must be open to activate buttons
-contains options that manipulate the general parameters of a country file
-name and acronym
-currencies used for parameters in the system and output
-which datasets are available and their characteristics
-Adding and deleting systems
-Viewing options:
-full spine vs. single policy
-search and replace
-formatting
-bookmarks
-More advanced (import/ export systems, add-ons etc.)
30
Systems Settings
set
exchange
rate
income used
for head
definition
Parameter
s’ currency
output currency
31
Database settings
characteristics of
dataset to be filled in
adding/ deleting
folder where micro-data stored
if different from default
32
Ribbon bar
Control display and formatting
•
•
•
•
View full spine vs. single policy
Conditional formatting (similar idea to EXCEL)
Bookmarks
Matrix view of income lists
33
Ribbon bar
Administration tools:
-adding and deleting countries
-accessing and administering the variables file
-updating progress: overview of available policy systems and datasets
More advanced & not covered in this course:
-available add-ons
-applications (EXCEL based)
34
EUROMOD Help
browse search
35
Context menus
36
Context Menus
• Activated by right-clicking
– Column headings
– Row headings
– Function headings/ parameter names
– Comments
• Intuitive options controlling the respective elements
37
IntelliSense
38
Variable Administration
name
List of all variables
existing in all
countries in
alphabetical order
Set vbl to monetary or non-monetary
automatic label
Description of variable for
countries where it is used
39
Adding a variable
new empty row;
fill in name and monetary
40
Naming a variable
acronyms
UI checks validity of name and
existence of the variable
41
Filtering variables
select filters
42
Running EUROMOD
select countries
select datasets
select systems
43
Running EUROMOD
extra options
44
Running EUROMOD
control display of run
log and error log
status
data and
systems running
run/ error log
run dialog
45
Output files
• micro-data (with an optional header)
Header
(optional)
• separate header file (optional)
Header
Detailed
run-time
(optional)
46
Output files
• Content manipulated in policy output_std_cc
• Usually including:
– All variables present in the input microdata file
– Simulated variables (i.e. simulated taxes and benefits)
– Standardized income lists
– (optional) non-standard income lists
– (optional) temporary variables
– (optional) Tax unit identification info
• Control level at which info is outputted (ex: individual, household etc.)
47
Summary Statistics Tool
•
•
•
•
Output of EUROMOD= micro-data
Process using a statistical software package (ex. Stata)
Only for training purposes- Summary Statistics Tool
Computes a range of commonly used indicators and statistics:
– poverty rates for the overall population and for selected groups and
the Gini coefficient
– distribution of household income, taxes and benefits by income group
– demographic information on households by income group
• Currently in Excel
• Computed indicators are fixed and cannot be changed  not for ‘real’
analysis!!!
• 7 tables produced in Excel
48
Summary statistics tool
folder where your output file is stored
49
Summary Statistics Tool
country and system on which
statistics calculated
6 tables with ‘fixed’
statistics
one sheet per output file
50
Implementing a simple reform
• Where:
– Simpleland
• What:
– make the child benefit more generous
• How:
– Open Simpleland
– Add a new system where your reform will be implemented
51
Implementing a simple reform
• How:
– Open the child benefit policy
– Make the changes in the new (reform) system
– Run EUROMOD
– Analyze results with the Summary Statistics Tool
52
Error handling
produce an error
53
Error handling
info on nature and location of the error
54
Error handling
•
•
•
•
Output folder-error log file (text format)
Same info as in the running dialog box
Error logs contain time stamp of their creation
Info about EUROMOD version, policy system where error occurred and
dataset used
55
Documentation
• MANUALS
– Euromod Terminology
– Running Euromod and Basic Concepts
– Euromod Functions
all in built-in help
• COUNTRY REPORTS (CR)
(https://www.iser.essex.ac.uk/euromod/resources-foreuromod-users/country-reports)
• DATA DESCRIPTION DOCUMENTS (DRD)
• WORKING PAPERS
(https://www.iser.essex.ac.uk/euromod/working-papers)
56
Country report
1.
2.
3.
4.
Basic information
– background information
– brief description of all policies
Simulation of taxes and benefits in EUROMOD
– scope and order of simulation
– detailed information on simulated policies (incl. assumptions)
Data
– general description and references to original data documentation
– data adjustment, imputations and assumptions
Validation
– policy validation
– income distribution validation: poverty and inequality
– “health warnings”
57
End of session
58
Exercise 1
• Producing summary statistics
59
EUROMOD functions and parameters
60
EUROMOD functions
• Building blocks to implement policies
– Parameters stored in XML and manipulated via the UI
– Calculation in EUROMOD executable (C++ code)
• Standardised simulation language
– Flexibility
– Harmonisation
– Parameterisation
– Consistency (e.g., errors handling)
– Sufficiency (any country any policy)
• Transparent and documented
– In-built HELP
– EMM_Functions manual (same info)
61
Structure of a function
Function name/ type
Parameter names
Switch: on/off/ toggle
Parameter values
Parameters are either compulsory or optional
62
Policies
• = block of functions that complete a ‘real’ policy simulation
• can be manipulated independently
– switch –affects all functions in the policy
– same policy may be repeated by simple referencing
– can be copied / moved
• order of policies is called ‘spine’
• policy names end (usually) with the country acronym
• each policy will have some explanation on what it is intended to simulate
in the comment columns
• policies can have any name
• ...but in practice we use some conventions
• can be:
– common to all countries (ex: defining uprating factors)
– country specific (ex: means-tested child benefit for single parents)
63
Policies
policy switch
policy name
policy description
social insurance contribution
policy made up of 3 functions
64
Manipulating functions and policies
right-click on
policy name to
activate menu
right-click on
function/
parameter names
to activate menu
65
Type of functions
• System functions
– functions used to define some general settings that are common
across countries (ex: uprating, default values for datasets etc.)
• Policy functions
– functions used to implement tax-benefit policies
• Special functions
– more advanced functions that perform more complicated tasks (loops,
changing parameters at run-time etc.)
– not covered in this course
66
Policy functions
Elig, ArithOp
67
Function Elig
•
•
implements conditions
sets a variable (by default sel_s) to 0 or 1, based on the condition in elig_cond.
Subsequent functions use this information via parameter who_must_be_elig
68
Function ArithOp
• Arithmetical calculator. The result of the parameter formula is stored as
output variable
69
Parameters
70
Parameters
• May be:
– Common to several functions
– Specific to one function
• May be:
– Compulsory (i.e. error generated if not used)
– Optional
– Which parameters are compulsory/ optional depends on the function
• Order of parameters in a function is not important
– (but order of functions in a policy is!!!!!!)
• Manipulated via context menu
– Only relevant parameters for the given function are shown
• Drag & drop can be used
71
Parameters
72
Common “eligibility” parameters (1)
– who_must_be_elig: function’s calculations are carried out if…
–
–
–
–
–
one_member (or one): one member of the assessment unit is eligible
one_adult: one adult member of the assessment unit is eligible
all_members (or all or taxunit): all members of the assessment unit are eligible
all_adults: all adult members of the assessment unit are eligible
nobody: calculations are carried out for each assessment unit (default)
• “eligibility” is determined by the variable indicated by the parameter
elig_var (by default sel_s)
–
–
0: person is not eligible
1: person is eligible
73
Common “eligibility” parameters (2)
74
Common “eligibility” parameters (3)
nobody
all_adults
all
one_adult
one
sel_s
dag
idhh
idperson
who_must_be_elig
1
11
80
1
1
1
0
0
1
1
12
60
0
1
1
0
0
1
1
13
40
0
1
1
0
0
1
2
21
80
1
1
1
0
1
1
2
22
6
0
1
1
0
1
1
3
31
80
1
1
1
1
1
1
4
41
40
0
0
0
0
0
1
4
42
40
0
0
0
0
0
1
75
Common output parameters
– Output_var, output_add_var, result_var
– func_Elig  sel_s
– Either output_var or output_add_var must be indicated
76
Common parameter TAX_UNIT
• TAX_UNIT allows for the definition of the assessment unit a function
refers to:
– Individuals
– Various definitions of family units
– Household units
• Compulsory for most policy functions
77
Common “limiting” parameters
– Lowlim (lower limit)
– Uplim (upper limit)
– Threshold (threshold)
78
Linking functions
79
Interactions between functions (1)
– Input: one function calculates a variable, which is used as an input by
a subsequent function.
80
Interactions between functions (2)
• The functions interact in three ways (+ replacement):
– Condition: one function (usually function Elig) evaluates a condition
and a subsequent function operates on the basis of the result of this
evaluation
81
Interactions between functions (3)
– Addition: one function calculates a part of a policy and a subsequent
function calculates another part of the policy and therefore needs to
add to the first part.
82
Result of a function
• It is always assigned to the head of the assessment unit
• For all other members of the unit and for those in not eligible units
(defined by who_must_be_elig) :
– output_var is set to zero.
– output_add_var not changed or set to 0 if undefined before
– result_var is set to zero.
83
Exercise 2
• Reforming child benefit in Estonia
84
End of session
85
InGRID Winter School in Cross-country
Microsimulation
Day 2
Silvia Avram, Chrysa Leventi, Iva Tasseva,
Alberto Tumino
18-20 February 2015
Course based on EUROMOD v. G2.0+
86
Outline: Day 2
• Morning (9:30-12:30 with a break at 11:00)
– EUROMOD Functions and Parameters (cont.)
– Hands-on practice: Exercises 3
– EUROMOD Functions and Parameters (cont.)
– Income Lists
• Afternoon (14:00-17:00 with a break at 15:30)
– Hands-on practice: Exercises 4 & 5
– Hands-on practice: Exercise 6
87
System functions
Uprate, SetDefault,DefOutput,
DefConst, DefVar
88
Function Uprate (I)
• Indices to uprate monetary variables to price level of policy year
name of
variable to
be uprated
value of
uprating factor
define factor to be
used later on
89
Function Uprate (II)
Aggregate variables
Using different values for different groups
90
Function SetDefault
• sets alternative values or variables if a dataset variable is missing
91
Function DefOutput
•
Determines the content of the output file
•
TAX_UNIT: level of aggregation
92
Function DefConst
•
To set up constants ... name them always starting with a $
93
Function DefVar
•
To set intermediate (temporary) variables not included in VarConfig.xml
94
Parameter values
95
Amount parameters
•
Monetary (numbers; use . for decimal) followed by their period:
•
– #m for monthly (no conversion)
– #y for yearly
– #q for quarterly
– #w for weekly
– #d for daily
– #l for labour day
– #s for six day labour week
Default is #m (monthly)
96
Query parameters (1)
•
•
•
frequently used ready made calculations
The results of a query is either yes/no or some (monetary or non monetary) value.
Well-documented in Help
97
Query parameters (2)
• Use IntelliSense to enter values
query symbol
98
Formula parameters
• Operations: ^, <min>, <max>, <abs>, (), !(), %, , /, *, \, +, -,
• Operands :
– variables (monetary and non monetary),
– incomelists
– queries
99
Condition parameters
•
•
•
logical and comparison operations to evaluate a condition with a yes/no result.
Conditions in by curly brackets {}, can be grouped by parenthesis ()
Negative condition (i.e. !) can be used with a single condition only
100
Footnote parameters
• They serve the further specification of other parameters. They are
identified by #i (i=number from 1 to....)
–
–
–
–
Limits
Amounts
Assessment units
Specification of queries
101
Parameter values and the assessment unit
level of interpretation
condition parameters
other parameters
monetary variables and
incomelists
assessment unit
assessment unit
non-monetary variables
and individual level
queries
individual
head of assessment unit
non individual level
queries
check manual
check manual
102
Exercise 3
• Reforming the child benefit in the UK
103
Policy functions
BenCalc, SchedCalc, Allocate
104
Function BenCalc (1)
•
•
•
Benefit calculator, used to implement a wide range of policy instruments, in
particular benefits
It combines the functionalities of the functions func_Elig and func_ArithOp
Calculates a sum of “components”, where the value of a component is only added
if a certain condition is fulfilled by at least one member of the assessment unit
105
Function BenCalc (2)
• compi_cond = func_Elig |elig_cond
• compi_perTU or compi_perElig = func_ArithOp | formula
– compi_perTU amount is added once
– compi_perElig amount is added once per individual fulfilling the
condition is added.
• either perTU OR perElig in one component
• can set upper and lower limits for earch component
– compi_lowlim
– compi_uplim
106
Function BenCalc (3)
• Withdraw parameters- subtract something from the calculated sum of
components
– withdraw_base: what is being subtracted (ex: variable, income list)
– withdraw_rate: what percentage of the base is being subtracted
– withdraw_start: set a minimum level of the sum of components
before any subtraction begins
– withdraw_end: level of the base where sum of the componentsbase*rate is 0
• Negative result automatically set to 0
• Result=max(Sum of components-max(BASE-START, 0)*RATE, 0)
• Rate and end cannot be used simultaneously
– If withdraw end is specified:
• RATE=(sum of components)/(END-START)
107
Function BenCalc (4)
108
Function SchedCalc (1)
• Used (mainly) for progressive taxes
• Tax schedule
– Tax bands: bandi_upLim / bandi_lowlim
– Tax rate: bandi_rate
– Tax base : base
• Instead rates, for fixed amounts use bandi_amount
109
Function SchedCalc (2)
•
•
Joint taxation: quotient
Result= ((Base/Quotient)*Tax schedule)*Quotient
110
Function SchedCalc (3)
• simple_prog: apply highest marginal tax rate reached by base on the
whole income
111
Function Allocate (1)
• default: result is assigned to the head of the assessment unit.
• Allocate reallocates amounts between members of assessment units
(subject to conditions)
112
Function Allocate (2)
• Split the amount of a variable
– share: which variable to split
– Amount to split first summed up across assessment unit members
– share_between: condition parameter; who are the members
‘participating’ in the split
– Default is all members of the assessment unit
– share_prop: in what proportion to split between the various qualifying
members (i.e. those satisfying the share_between condition)
– Default is sharing in equal proportions
113
Special functions
114
Special functions (advanced)
• func_AddOn_xxx (implement extended functionalities not part of the standard taxbenefit calculations, such Effective Marginal Tax Rates, different budget sets)
•
•
•
•
•
func_Loop and func_UnitLoop (repeat part (or all) of the tax-benefit calculations)
func_Store and func_Restore (set a variable to the initial (or other previous) value
func_ChangeParam (modify parameters during the model run)
func_Totals (calculates aggregates over groups of persons/households)
func_DropUnit and func_KeepUnit (drops certain persons/households from the
calculations)
• func_ILVarOp (performs operations on variables that are part of an income list)
• func_RandSeed (generates random numbers)
•
func_CallProgramme (calls another programme, e.g., Stata)
115
Incomelists
116
Incomelists
• Aggregates of several variables
• Standardised output (e.g. ils_dispy)
• func_DefIL : special policy (i.e. ILDef_cc) or any other policy
• Once defined it is available for all subsequent functions and policies
• Naming convention: prefix il_ for “normal”, ils_ for “standard”
• No tax unit defined all income lists built at the individual level
• Behave like a monetary variable
117
Function DefIL (1)
• Components:
─Variables
─Pre-defined income lists
─Fixed amounts
─Constants
•Operations:
─+, ─fractions can be used
118
Function DefIL (2)
• If you want to take out a component in a specific policy system
replace operation with n/a
119
Standardized income lists
• Defined in every country
• Built in a comparable way to facilitate cross-national analysis
– ils_earns: earnings
– ils_origy: market incomes
– ils_pen: public pensions
– ils_bennt: non-means-tested benefits
– ils_benmt: means-tested benefits
– ils_ben: all benefits and public pensions
– ils_tax: taxes
– ils_sicee: employee SICs
– ils_sicse: self-employed SICs
– ils_sicer: employer SICs
– ils_sicct: contributed SICs
– ils_dispy: disposable income
– ils_bensim: simulated benefits
– ils_taxsim: simulated taxes
120
Incomelists – Matrix view
•
Summary of income list components
121
End of session
122
Exercise 4
• Reforming the means-tested income support for families with children in
compulsory education in Greece
123
Exercise 5
• Reforming the health tax in Denmark
124
Exercise 6
• Reforming the social assistance (and income tax ) in Bulgaria
125
End of session
126
InGRID Winter School in Cross-country
Microsimulation
Day 3
Silvia Avram, Chrysa Leventi, Iva Tasseva,
Alberto Tumino
18-20 February 2015
Course based on EUROMOD v. G2.0+
127
Outline: Day 3
• Morning (9:30-12:30 with a break at 11:00)
– Assessment (tax) units in EUROMOD
– Hands-on practice: Exercise 7
• Afternoon (14:00-17:00 with a break at 15:30)
– Hands-on practice: Exercises 8 & 9
– Presentations by participants
– Using EUROMOD after the course
– Q&A
128
Assessment (Tax) Unit
129
Assessment Unit
•
•
•
•
Unit: group of household members to be considered together
Function DefTU, used at sheet TUDef_cc or anywhere
Defined the first time it is used by the model (see func_UpdateTU)
Name convention: tu_xxxxxx_cc defined
130
Types of assessment units
Parameter type: defines the composition of the tax unit
• HH: all individuals of the household are in the same unit.
• IND: each individual of the household forms its own unit.
• SUBGROUP: individuals determined by parameter members form an unit. The
household may be split into several units of different size.
• Micro data used by EUROMOD─ sample of households
─ all individuals in a selected household
─if assessment unit includes individuals outside the household cannot be
reconstructed exactly
Subgroup
Individual
Household
131
Some examples of assessment units
description
typical family
couple.without
children
lone parent
single
two singles
living together
large family
idhh idperson idpartner idmother idfather dag Household Individual
family
1
1
1
1
2
2
3
3
4
5
5
101
102
103
104
201
202
301
302
401
501
502
102
101
0
0
202
201
0
0
0
0
0
0
0
102
102
0
0
0
301
0
0
0
0
0
101
101
0
0
0
0
0
0
0
30
28
3
1
56
55
35
6
25
22
23
A
A
A
A
A
A
A
A
A
A
A
A
B
C
D
A
B
A
B
A
A
B
A
A
A
A
A
A
A
A
A
A
B
6
6
6
6
6
6
601
602
603
604
605
606
602
601
0
0
0
0
606
0
602
602
602
0
0
0
601
601
601
0
48
45
20
15
10
70
A
A
A
A
A
A
A
B
C
D
E
F
A
A
A
A
A
B
132
Head of a tax unit
•
•
The head of a tax unit is
– the richest member of the unit (System Configuration: Income for Head
Definition)
– the oldest
– the lowest idperson
ExtHeadCond: further conditions (e.g., female)
parameter name
compulsory /
optional
optional
default value
description
ils_origy
ExtHeadCond *
value
type
variable
or
incomelist
condition
optional
!{IsDepChild}
StopfIfNoHeadFound
yes/no
optional
no
NoChildIfHead
yes/no
optional
no
NoChildIfPartner
yes/no
optional
no
variable or incomelist used for determining who is
the richest person in the assessment unit, see
description of parameter ExtHeadCond
condition further defining the head of the
assessment unit
the condition is &-linked with the following
hardwired head condition:
{HeadDefInc>anyother:HeadDefInc} |
({HeadDefInc>=anyother:HeadDefInc} &
{dag>anyother:dag}) |
({HeadDefInc>=anyother:HeadDefInc} &
{dag>=anyother:dag} &
{idperson<anyother:idperson})
yes: error is issued if ExtHeadCond rules out all
household members
no: no error issued, ExtHeadCond dropped for
affected households
if yes (possible) child status is removed if person is
head
if yes (possible) child status is removed if person is
partner
HeadDefInc
133
Members of a tax unit (1)
•
•
members usually defines relations with respect to the head of the unit
possible values
– Partner
– Children (Own/ Loose, dependent...etc)
– Dependent parents
– Dependent Relatives
•
status of each member (i.e. Partner, OwnDepChild…) is defined by a xxxCond
parameter
– PartnerCond
– OwnChildCond
– DepChildCond
– OwnDepChildCond
– LooseDepChildCond
– DepParentCond
– DepRelativeCond
134
Members of a tax unit (2)
• Condition parameters and default values fully documented in Help
parameter name
value
compulsory
type
/ optional
categorical compulsory if
type=
default value
description
n/a
PartnerCond *
DepChildCond *
condition
condition
optional
optional
OwnChildCond *
OwnDepChildCond *
LooseDepChildCond
*
DepParentCond *
DepRelativeCond *
LoneParentCond *
condition
condition
condition
optional
optional
optional
condition
condition
condition
optional
optional
optional
{head:idperson=idpartner}
{0}
{Default}=!{isparent}&{idpartner<=0}
see **
{isownchild}&{isdepchild}
{idmother=0}&{idfather=0}&{isdepchild}
respectively {idparent=0}&{isdepchild}
see ***
{0}
{isparentofdepchild}&{idpartner<=0}
defines which members of the household form a unit if
type=SOUBGROUP
syntax: status type & status type & status type ...
where status type can take the values:
Partner: defined by parameter PartnerCond
OwnDepChild: defined by OwnDepChildCond
LooseDepChild: defined by LooseDepChildCond
OwnChild: defined by OwnChildCond
DepParent: defined by DepParentCond
DepRelative: defined by DepRelativeCond
note, that the head is obviously always part of unit and
(usually) relations are defined with reference to the head
condition defining who is a partner
condition defining who is a dependent child
members
SOUBGROUP
condition defining who is an own child
condition defining who is an own dependent child
condition defining who is a loose dependent child
condition defining who is a dependent parent
condition defining who is a dependent relative
condition defining who is a lone parent
* variables may be used with the prefixes “head:” or “partner:”.{Default} can be used to further define default condition (see section 14.4)
** OwnChildCond: {head:idperson=idmother}|{head:idperson=idfather}|{partner:idperson=idmother}|{partner:idperson=idfather}
respectively: {head:idperson=idparent}|{partner:idperson=idparent}
*** DepParentCond: {head:idmother=idperson}|{head:idfather=idperson}|{partner:idmother=idperson}|{partner:idfather=idperson}
respectively: ({head:idparent=idperson}|{partner:idparent=idperson})|({idpartner>0}&({head:idparent=idpartner}|{partner:idparent=idpartner}))
135
Members of a tax unit (3)
• head: subsequent variable refers to the head of the unit
• partner: subsequent variable refers to the partner of the head of the unit
• {default} default setting, can be combined with further specifications
136
Members of a tax unit (4)
• DepChildCond: determines who is dependent children
• OwnDepChild: “a son or daughter”. See OwnChildCond
• LooseDepChild: “someone who is depend child but doesn’t cohabit with
parent/s”
idperson
101
102
103
104
105
106
idpartner
102
101
0
0
0
0
idmother
0
0
102
102
102
102
idfather
0
0
101
101
101
101
dag
44
40
21
19
17
10
IsInEducation
no
no
no
no
yes
yes
ils_origy
2500
1200
1000
800
0
0
IsDepChild
0
0
0
0
1
1
assessment unit
A
A
B
C
A
A
137
Avoiding to split up families
parameter name
AssignDepChOfDependents
value
type
yes/no
AssignPartnerOfDependents yes/no
compulsory /
optional
optional
default
value
no
optional
no
description
if yes dependent children of dependent unit members (i.e.
persons who are not head or partner of the unit) are assigned to
the unit
child/parent relation is identified by variables idmother, idfather
respectively idparent
if yes partners of dependent unit members (i.e. persons who are
not head or partner of the unit) are assigned to the unit
partner relation is identified by variable idpartner
138
Using conditions which refer to income
•
If the assessment unit is bigger than the individual, the level of interpreting
monetary variables or income lists must be considered carefully (using
footnotes)
•
Queries can be used to define income conditions (e.g. GetParentIncome,
GetCoupleIncome, …)
139
Updating assessment units
•
•
•
The tax unit is defined/ calculated the first time it is used by the model:
household members are assigned to respective units once an assessment unit is
first used.
This assignment is not changed with subsequent uses, even if circumstances
change.
However, the reassessment of the units can be enforced by using the function
UpdateTU.
140
102
0
0
101
0
0
0 102 101
0 102 101
202
0
0
201
0
0
0 202 201
0 202 201
5,202 5,206 5,205
5,201
0
0
0 5,202 5,201
0 5,202 5,201
5,206
0
0
5,205
0
0
0
0
0
0
0 9,201
0
0 9,201
0
0 9,202
65
60
30
28
29
25
3
2
40
38
10
15
70
70
80
38
34
11
0
0
0
147
1,007
891
0
0
1,831
0
0
0
0
0
0
3,502
2,324
0
tu_tinfajt_headid
tu_tinfajt_ispartner
tu_tinfajt_isdepchild
tu_tinfajt_isdepparent
tu_tinfajt_isloneparent
101
102
103
104
201
202
203
204
5,201
5,202
5,203
5,204
5,205
5,206
9,201
9,202
9,203
9,204
il_tinty
#1_level
AssignDepChOfDependen
yes
ts
AssignPartnerOfDepende
yes
nts
{default} & !{IsMarried} &
LoneParentCond
{nDepChOfPerson > 0}
1
1
1
1
2
2
2
2
52
52
52
52
52
52
92
92
92
92
dag
DepParentCond
idfather
DepChildCond
idmother
PartnerCond
idpartner
Members
on
tu_tinfajt
SUBGROUP
Partner & OwnDepChild
& DepParent
{Default} & {IsMarried}
{Default} & {dag<25} &
{il_tinty#1<=8000#y}
{Default} & {dag>65} &
{il_tinty#1<=8000#y}
tu_individual_es
idperson
func_DefTu
Name
Type
idhh
Tax unit in output file
101
101
103
104
201
201
201
201
5,201
5,201
5,201
5,201
5,201
5,201
9,202
9,202
9,203
9,202
0
1
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
1
1
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
141
Exercise 7
• Reforming the UK Child Benefit
142
End of session
143
Exercise 8
• Introducing a benefit cap in the UK
144
Exercise 9
• Introducing the Belgian social insurance contributions for pensioners &
survival pensioners in the UK
Using EUROMOD after the course
146
Model access
• Web https://www.iser.essex.ac.uk/euromod
– Statistics on the distribution and decomposition of disposable income
– Country Reports
– Working Papers
• Model is freely available for non-commercial use
– contact euromod@essex.ac.uk to obtain the link for downloading
(incl. manuals)
147
Data Access conditions
Data access is subject to conditions set by the original data provider
• EU-SILC UBD
– Access depends on being in an institution recognised by Eurostat as
a “research entity”. For more info see
http://epp.eurostat.ec.europa.eu/portal/page/portal/microdata/doc
uments/How_to_apply_for_microdata_access.pdf
– In a second stage you need to be part of a “research proposal”
accepted by Eurostat that includes the use of EUROMOD
– When you are ready to submit the second stage proposal contact
euromod@essex.ac.uk
– Allow at least 4 months for the whole process
• Other data: relatively straightforward procedures
148
Responsibilities of EUROMOD hand-on users
•
•
•
•
•
•
Respect data access rules and conditions
Acknowledge EUROMOD when it is used
Submit all papers using EUROMOD for inclusion in the WP series
Take responsibility for your own use of the model
Tell us about bugs or errors
Keep us informed about what you are working on and when you are
working actively: that way we can keep you informed of relevant changes
149
•Q&A
150
End of session