InGRID Winter School in Cross-country Microsimulation

Day 1

Silvia Avram, Chrysa Leventi, Iva Tasseva, Alberto Tumino

ISER, U Essex18-20 February 2015

Course based on EUROMOD v. G2.0+

1

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

2

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

3

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

4

Tax-benefit microsimulation and EUROMOD

5

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.

6

Tax-benefit models

• Micro: household micro-data

• Simulation: taxes and benefits

• Contribution: impact on disposable income

7

Why tax-benefit microsimulation?0

.25

.5.7

5m

argi

nal t

ax r

ate

1 2 3 4 5 6 7 8 9 10deciles

Source: Euromod using EU-SILC 2004

• Population diversity and frequency

• Policy complexity: detail and interactions

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

9

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.)

10

Examples-Counterfactual scenarios (I)• To what extent tax-benefit systems support those who became

unemployed at the onset of the Great Recession?

11

Source: Fernandez et al. (2013) Review of Income and Wealth

Examples-Counterfactual scenarios (II)• Poverty effects of child related policy instruments in Poland

12

0%

5%

10%

15%

20%

25%

30%

35%

FGT0 FGT1 FGT2

Baseline 2005

PL-2007

Austrian system

French system

UK system

No fam ben

Source: Levy, Morawski and Myck, Euromod Working Paper 3/08

Examples-Work incentives• Calculate Marginal Effective Tax Rates

13

Source: Jara and Tumino (2013) International Journal of Microsimulation

Examples-EU-wide reforms• Net replacement rates with and without a EU-wide UI benefit

14

2030405060708090

100110

Nati

onal

EMU-

UI fl

atEM

U-U

I %N

ation

alEM

U-U

I flat

EMU-

UI %

Nati

onal

EMU-

UI fl

atEM

U-U

I %N

ation

alEM

U-U

I flat

EMU-

UI %

Nati

onal

EMU-

UI fl

atEM

U-U

I %N

ation

alEM

U-U

I flat

EMU-

UI %

Nati

onal

EMU-

UI fl

atEM

U-U

I %N

ation

alEM

U-U

I flat

EMU-

UI %

Nati

onal

EMU-

UI fl

atEM

U-U

I %N

ation

alEM

U-U

I flat

EMU-

UI %

DE EE EL ES FR IT LV AT PT FI

p25-p75 Median Mean

Source: Jara & Sutherland (2014) Social Situation Monitor Research Note 3/13

Examples-Behavioural reactions• Calculating labour supply elasticities

15

0

.2

.4

.6

.8

Own-wage elasticity

EE05

UK01

PL05

SW

01

FR01

FI98

PT01

US05

HU05

DK98

BE01

GE01

IE01

NL01

IT98

AT98

SP01

GR98

Married women

0

.2

.4

.6

.8

Own-wage elasticity

UK01

PT01

PL05

IT98

NL01

FR01

SW

01

AT98

SP01

US05

HU05

EE05

FI98

GR98

BE01

GE01

DK98

IE01

Married men

0

.2

.4

.6

.8

Own-wage elasticity

PT01

NL01

PL05

AT98

FR01

HU05

EE05

GR98

SW

01

US05

GE01

UK01

IT98

DK98

BE01

FI98

SP01

IE01

Single men

0

.2

.4

.6

.8

Own-wage elasticity

PT01

HU05

PL05

EE05

FR01

DK98

AT98

NL01

GE01

SP01

FI98

SW

01

US05

UK01

IE01

GR98

BE01

IT98

Single women

Point estimate 95% confidence interval

Source: Bargain et al. (2013) Journal of Human Resources

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

16

Overview of EUROMOD design

17

Structure

18

Input microdata(text file)

Policy parameters(XML files-EUROMOD UI)

Simulations(EURMOD Engine in C++)

Output microdata with additional simulated variables (text file)

• 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

EUROMOD input dataset19

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 incomeyse: self-employment income

** further bb’s for additional information/detail eg. ysebs: business self-employment income

exception id*, eg. idperson, idmother

Variable name convention20

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

21

EUROMODUser Interface

22

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

23

EUROMOD folder structure

24

Linking EUROMOD to content files

25

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

26

User Interface (UI)

27

Run button

Ribbon bar

Country files

Working environment28

open country

policy systems

comments

policies

Ribbon bar29

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.)

Systems Settings30

set exchange rate

Parameters’ currency output currency

income used for head definition

Database settings31

adding/ deletingfolder where micro-data stored if different from default

characteristics of dataset to be filled in

Ribbon bar

32

Control display and formatting• View full spine vs. single policy• Conditional formatting (similar idea to EXCEL)• Bookmarks• Matrix view of income lists

Ribbon bar

33

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)

EUROMOD Help

34

browse search

Context menus

35

Context Menus

• Activated by right-clicking– Column headings– Row headings– Function headings/ parameter names– Comments

• Intuitive options controlling the respective elements

36

IntelliSense

37

Variable Administration

38

List of all variables existing in all countries in alphabetical ordername

Description of variable for countries where it is used

automatic label

Set vbl to monetary or non-monetary

Adding a variable

39

new empty row;fill in name and monetary

Naming a variable

40

acronyms

UI checks validity of name and existence of the variable

Filtering variables

41

select filters

Running EUROMOD

42

select countries

select systems

select datasets

Running EUROMOD43

extra options

Running EUROMOD44

run dialogdata and systems running

status

control display of run log and error log

run/ error log

Output files• micro-data (with an optional header)

• separate header file (optional)

Header

Detailed run-time (optional)

Header (optional)

45

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.)

46

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

47

Summary statistics tool48

folder where your output file is stored

Summary Statistics Tool

49

country and system on which statistics calculated

one sheet per output file

6 tables with ‘fixed’ statistics

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

50

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

51

Error handling

52

produce an error

Error handling53

info on nature and location of the error

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

54

Documentation

• MANUALS– Euromod Terminology all in built-in help– Running Euromod and Basic Concepts– Euromod Functions

• COUNTRY REPORTS (CR)(https://www.iser.essex.ac.uk/euromod/resources-for-euromod-

users/country-reports)• DATA DESCRIPTION DOCUMENTS (DRD)• WORKING PAPERS(https://www.iser.essex.ac.uk/euromod/working-papers)

55

Country report

1. Basic information– background information– brief description of all policies

2. Simulation of taxes and benefits in EUROMOD– scope and order of simulation– detailed information on simulated policies (incl. assumptions)

3. Data– general description and references to original data documentation– data adjustment, imputations and assumptions

4. Validation– policy validation – income distribution validation: poverty and inequality– “health warnings”

56

End of session

57

Exercise 1

• Producing summary statistics

58

EUROMOD functions and parameters

59

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)

60

Structure of a function

Parameters are either compulsory or optional

61

Function name/ type

Parameter names Parameter values

Switch: on/off/ toggle

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)

62

Policies

63

social insurance contribution policy made up of 3 functions

policy name

policy switch

policy description

Manipulating functions and policies

64

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

65

Policy functionsElig, ArithOp

66

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

67

Function ArithOp• Arithmetical calculator. The result of the parameter formula is stored as

output variable

68

Parameters

69

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

70

Parameters

71

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

72

Common “eligibility” parameters (2)

73

Common “eligibility” parameters (3)

        who_must_be_elig

idhhidperson

dag sel_s oneone_adult

allall_adults

nobody

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

74

Common output parameters

– Output_var, output_add_var, result_var– func_Elig sel_s

– Either output_var or output_add_var must be indicated

75

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

76

Common “limiting” parameters

– Lowlim (lower limit)– Uplim (upper limit)– Threshold (threshold)

77

Linking functions

78

Interactions between functions (1)

– Input: one function calculates a variable, which is used as an input by a subsequent function.

79

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

80

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.

81

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.

82

Exercise 2

• Reforming child benefit in Estonia

83

End of session

84

InGRID Winter School in Cross-country Microsimulation

Day 2

85

Silvia Avram, Chrysa Leventi, Iva Tasseva, Alberto Tumino

18-20 February 2015Course based on EUROMOD v. G2.0+

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

86

System functionsUprate, SetDefault,DefOutput,

DefConst, DefVar

87

88

Function Uprate (I) • Indices to uprate monetary variables to price level of policy year

88

name of variable to be uprated

value of uprating factor

define factor to be used later on

Function Uprate (II)89

Aggregate variables Using different values for different groups

90

Function SetDefault• sets alternative values or variables if a dataset variable is missing

90

91

Function DefOutput• Determines the content of the output file

• TAX_UNIT: level of aggregation

91

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

Parameter values

94

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)

95

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

96

Query parameters (2)• Use IntelliSense to enter values

97

query symbol

Formula parameters

• Operations: ^, <min>, <max>, <abs>, (), !(), %, , /, *, \, +, -,• Operands :

– variables (monetary and non monetary), – incomelists – queries

98

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

99

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

100

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

101

Exercise 3

• Reforming the child benefit in the UK

102

Policy functionsBenCalc, SchedCalc, Allocate

103

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

104

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

105

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 components-

base*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)

106

Function BenCalc (4)107

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

108

109

Function SchedCalc (2)

• Joint taxation: quotient• Result= ((Base/Quotient)*Tax schedule)*Quotient

109

Function SchedCalc (3)

• simple_prog: apply highest marginal tax rate reached by base on the whole income

110

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)

111

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

112

Special functions

113

Special functions (advanced)

• func_AddOn_xxx (implement extended functionalities not part of the standard tax-benefit 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)

114

Incomelists

115

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

116

Function DefIL (1)

117

• Components:─ Variables─ Pre-defined income lists─ Fixed amounts─ Constants

• Operations:─ +, -─ fractions can be used

Function DefIL (2)

118

• If you want to take out a component in a specific policy system replace operation with n/a

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

119

Incomelists – Matrix view • Summary of income list components

120

End of session

121

Exercise 4

• Reforming the means-tested income support for families with children in compulsory education in Greece

122

Exercise 5

• Reforming the health tax in Denmark

123

Exercise 6

• Reforming the social assistance (and income tax ) in Bulgaria

124

End of session

125

InGRID Winter School in Cross-country Microsimulation

Day 3

126

Silvia Avram, Chrysa Leventi, Iva Tasseva, Alberto Tumino

18-20 February 2015Course based on EUROMOD v. G2.0+

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

127

Assessment (Tax) Unit

128

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

129

130

Types of assessment unitsParameter 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

130

Individual HouseholdSubgroup

131

Some examples of assessment unitsdescription idhh idperson idpartner idmother idfather dag Household Individual family

typical family 1 101 102 0 0 30 A A A1 102 101 0 0 28 A B A1 103 0 102 101 3 A C A1 104 0 102 101 1 A D A

couple.without children

2 201 202 0 0 56 A A A2 202 201 0 0 55 A B A

lone parent 3 301 0 0 0 35 A A A3 302 0 301 0 6 A B A

single 4 401 0 0 0 25 A A Atwo singles living together

5 501 0 0 0 22 A A A5 502 0 0 0 23 A B B

large family 6 601 602 606 0 48 A A A6 602 601 0 0 45 A B A6 603 0 602 601 20 A C A6 604 0 602 601 15 A D A6 605 0 602 601 10 A E A6 606 0 0 0 70 A F B

131

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 value type

compulsory / optional

default value description

HeadDefInc variable or incomelist

optional ils_origy variable or incomelist used for determining who is the richest person in the assessment unit, see description of parameter ExtHeadCond

ExtHeadCond * condition optional !{IsDepChild} 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})

StopfI fNoHeadFound yes/no optional no yes: error is issued if ExtHeadCond rules out all household members no: no error issued, ExtHeadCond dropped for affected households

NoChildI fHead yes/no optional no if yes (possible) child status is removed if person is head

NoChildI fPartner yes/no optional no if yes (possible) child status is removed if person is partner

132

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

133

Members of a tax unit (2)• Condition parameters and default values fully documented in Help

134

parameter name value type

compulsory / optional

default value description

members categorical compulsory if type= SOUBGROUP

n/a 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

PartnerCond * condition optional {head:idperson=idpartner} condition defining who is a partner DepChildCond * condition optional {0}

{Default}=!{isparent}&{idpartner<=0} condition defining who is a dependent child

OwnChildCond * condition optional see ** condition defining who is an own child OwnDepChildCond * condition optional {isownchild}&{isdepchild} condition defining who is an own dependent child LooseDepChildCond *

condition optional {idmother=0}&{idfather=0}&{isdepchild} respectively {idparent=0}&{isdepchild}

condition defining who is a loose dependent child

DepParentCond * condition optional see *** condition defining who is a dependent parent DepRelativeCond * condition optional {0} condition defining who is a dependent relative LoneParentCond * condition optional {isparentofdepchild}&{idpartner<=0} 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

135

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 idpartner idmother idfather dag I sInEducation ils_origy I sDepChild assessment unit

101 102 0 0 44 no 2500 0 A 102 101 0 0 40 no 1200 0 A 103 0 102 101 21 no 1000 0 B 104 0 102 101 19 no 800 0 C 105 0 102 101 17 yes 0 1 A 106 0 102 101 10 yes 0 1 A

136

137

Avoiding to split up families

parameter name value type

compulsory / optional

default value

description

AssignDepChOfDependents yes/no optional no 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

AssignPartnerOfDependents yes/no optional no 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

137

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, …)

138

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.

139

Tax unit in output file

idhhidperson

idpartner

idmother

idfather dag

il_tinty

tu_tinfajt_headid

tu_tinfajt_ispartner

tu_tinfajt_isdepchild

tu_tinfajt_isdepparent

tu_tinfajt_isloneparent

1 101 102 0 0 65 0 101 0 0 0 01 102 101 0 0 60 0 101 1 0 0 01 103 0 102 101 30 0 103 0 0 0 01 104 0 102 101 28 147 104 0 0 0 02 201 202 0 0 29 1,007 201 0 0 0 02 202 201 0 0 25 891 201 1 0 0 02 203 0 202 201 3 0 201 0 1 0 02 204 0 202 201 2 0 201 0 1 0 0

52 5,201 5,202 5,206 5,205 40 1,831 5,201 0 0 0 052 5,202 5,201 0 0 38 0 5,201 1 0 0 052 5,203 0 5,202 5,201 10 0 5,201 0 1 0 052 5,204 0 5,202 5,201 15 0 5,201 0 1 0 052 5,205 5,206 0 0 70 0 5,201 0 0 1 052 5,206 5,205 0 0 70 0 5,201 0 0 1 092 9,201 0 0 0 80 0 9,202 0 0 1 092 9,202 0 0 9,201 38 3,502 9,202 0 0 0 192 9,203 0 0 9,201 34 2,324 9,203 0 0 0 092 9,204 0 0 9,202 11 0 9,202 0 1 0 0

func_DefTu  onName tu_tinfajtType SUBGROUP

MembersPartner & OwnDepChild & DepParent

PartnerCond {Default} & {IsMarried}

DepChildCond{Default} & {dag<25} & {il_tinty#1<=8000#y}

DepParentCond{Default} & {dag>65} & {il_tinty#1<=8000#y}

#1_level tu_individual_esAssignDepChOfDependents

yes

AssignPartnerOfDependents

yes

LoneParentCond{default} & !{IsMarried} & {nDepChOfPerson > 0}

140

Exercise 7

• Reforming the UK Child Benefit

141

End of session

142

Exercise 8

• Introducing a benefit cap in the UK

143

Exercise 9

• Introducing the Belgian social insurance contributions for pensioners & survival pensioners in the UK

144

Using EUROMOD after the course

• 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)

Model access

146

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/documents/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

Data Access conditions

147

• 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

Responsibilities of EUROMOD hand-on users

148

•Q&A

149

End of session

150