microsimulation in the uk: taxben · tax and benefit systems since 1975 – plus, crucially,...

© Institute for Fiscal Studies

Microsimulation in the UK: TAXBEN

Stuart Adam, IFS

Workshop on microsimulation modelling for fiscal policy analysis

European Commission Joint Research Centre, Seville

23 September 2016


TAXBEN

• Static microsimulation model of the UK tax and benefit system

• First built early 1980s, continuous development since

– Now increasing investment in upgrades

• Current version written in Delphi (Object Pascal)

• Not open source, and poorly documented (for now)


Coverage

Tax and benefit systems since 1975

– Plus, crucially, hypothetical alternatives

Includes:

• Income tax

• Employee and employer social security contributions

• Council tax (local housing tax)

• VAT

• Most taxes on specific goods (fuel, alcohol, tobacco)

• State pensions and all significant state benefits and tax credits

But not:

• ‘Capital taxes’: capital gains tax, inheritance tax, stamp duties (property and share transaction taxes)

• ‘Business taxes’: corporation tax, business rates (property tax)

• Public services


Data

• Runs on several household survey datasets (with population weights)

– Family Resources Survey (from 1994)

– Living Costs and Food Survey (from 1978)

– English Longitudinal Study of Ageing (from 2002)

– (Labour Force Survey 2007)

• Hoping to add more soon

– Wealth and Assets Survey

– Understanding Society

• Also simulated datasets (e.g. simulated lifecycles)

• Not administrative data

– Doesn’t contain crucial information (family circumstances, rent, etc.)

– Sometimes use it separately


Static model

• Single period

• No behavioural responses or economic effects embedded in model

• But used extensively as input to behavioural modelling

– e.g. TAXBEN-generated budget constraints used to estimate labour supply models and simulate labour supply under different tax and benefit systems

– also consumer demand systems, etc.

• Allows us to use different behavioural models as appropriate

– Best model to use won’t be the same in every case

– TAXBEN calculator can be called by other programs

• Recent development: FORTAX

– Slimmed-down tax and benefit model written in Fortran

– Less functionality than TAXBEN, but faster

– Useful for dynamic lifecycle models of behaviour (need millions of runs)


Main UK microsimulation models outside IFS

• Static tax and benefit models

– IGOTM (HM Treasury) and PSM (Department for Work and Pensions)

– EUROMOD (based at University of Essex)

– One or two think tanks

• Dynamic models with a tax/benefit element

– PENSIM2 (Department for Work and Pensions)

– SAGE (LSE)

– LINDA (developed by NIESR, occasionally used by HM Treasury)


The Institute for Fiscal Studies

• An independent not-for-profit microeconomic research institute

• Aim: ‘to promote the development of effective policy on taxation and government spending by providing high-quality impartial evidence and analysis to inform the public debate’

– Bridge the gap between academia and policy discussion

• About 50 research staff, of whom about 6 main TAXBEN users

– Plus network of academic Research Fellows etc.

• Funded by research grants

• Unique and important role in UK policy debate


How is TAXBEN used?

• Best known for estimating revenue and distributional effects

• Also analysis of work incentives...

– Interesting in its own right

• ...which feeds into labour supply modelling

– Estimating/simulating effects of policies on employment, hours of work, earnings

• Input to other behavioural modelling too

– Benefit take-up, consumer demand, human capital investment, etc.


When is TAXBEN used?

Reactive:

• Day after Budgets etc.

• Election campaigns (government’s record, parties’ proposals)

• Ad hoc analysis of new announcements/proposals

Proactive:

• Annual IFS Green Budget (analysis of options for Budget)

• Academic / policy research projects

– ‘The trade-off between work incentives and redistribution’

– ‘A retrospective evaluation of the EU VAT system’

– ‘The effect of working families’ tax credit on lone mothers’ labour supply’

– The Mirrlees Review of taxation

Increasing the income tax personal allowance Distributional impact of an increase from £10,000 to £12,500


0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

Poorest 2 3 4 5 6 7 8 9 Richest All

Ch

an

ge

in

ne

t in

com

e

Household net income decile group Assumes higher-rate threshold held constant.

Source: Figure 7.4 of The IFS Green Budget: February 2014

Impact of tax and benefit reforms between January 2010 and April 2019

Source: IFS post-Budget briefing, July 2015

-14%

-12%

-10%

-8%

-6%

-4%

-2%

0%

-£7,000

-£6,000

-£5,000

-£4,000

-£3,000

-£2,000

-£1,000

£0


Income decile group

£ per year (left axis)

% of net income (right axis)



In cross-section, increase in out-of-work benefits is most progressive

0

.01

.02

.03

.04

.05

Pro

po

rtio

na

l ch

an

ge

in

ne

t in

com

e


Net income decile

Out-of-work benefits In-work benefits Income tax


Over the lifetime, increases in in-work and out-of-work benefits are similarly progressive

0

.00

2.0

04

.00

6.0

08

Pro

po

rtio

na

l ch

an

ge

in

ne

t in

com

e


Net income decile

Out-of-work benefits In-work benefits Income tax


Explanation: the poorest individuals spend most of working-age life in work

0.2

.4.6

.81

Pro

po

rtio

n e

mp

loy

ed


Cross-section Lifetime

Effect of tax and benefit reforms on income inequality


-0.04

-0.03

-0.02

-0.01

0.00

0.01

0.02

0.03

0.04

0.05

1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2009

Ch

an

ge

in

Gin

i

Uprated in line with RPI

Uprated in line with GDP

Taxes RPI-uprated, benefits GDP-uprated

Source: Adam and Browne (2010)


Universal credit and relative poverty rates

10%

15%

20%

25%

2010 2011 2012 2013 2014 2015 2016

Child poverty (BHC) Working-age non-parent poverty (BHC)

Without UC Without UC

With UC With UC

Note: take-up assumptions important here

Source: J. Browne, A. Hood and R. Joyce (2013), Child and working-age poverty in

Northern Ireland from 2010 to 2020


Effect of universal credit on work incentives

Percentage point change in average:

RR PTR EMTR

Single, no children –0.9 –1.5 +0.7

Lone parent +0.3 +3.6 –5.2

Partner not working, no children –3.2 –3.4 –0.4

Partner not working, children –5.7 –10.7 +1.2

Partner working, no children +0.1 +0.1 –0.3

Partner working, children +0.9 +2.5 +0.0

All –0.7 –0.7 –0.1

UC gets rid of many of the very weakest work incentives:

– reduces number of people with PTRs >75% by half (1.5m)

– reduces number of people with EMTRs >85% by more than 90% (0.5m)

Effect on average work incentives:



Effect of UC on average PTR, by earnings

-30

-25

-20

-15

-10

-5

0

5

10

15

£0 £10,000 £20,000 £30,000 £40,000 £50,000

Ch

an

ge

in

me

an

PT

R (

pp

ts)

Labour cost

All Single, no children

Lone parent Partner not working, no children

Partner not working, children Partner working, no children

Partner working, children


Distribution of VAT payments in the UK


€0

€20

€40

€60

€80

€100

€120

€140

€160

€180

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

Poorest 2 3 4 5 6 7 8 9 Richest

Income Decile Group

% of income % of expenditure Cash amount (€/week, right axis)

Source: IFS et al, ‘A retrospective evaluation of the EU VAT system’

Distribution of VAT payments in the UK


€0

€20

€40

€60

€80

€100

€120

€140

€160

€180

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

Poorest 2 3 4 5 6 7 8 9 Richest

Expenditure Decile Group

% of income % of expenditure Cash amount (€/week, right axis)


Revenue-neutral move to a uniform VAT rate Effect on spending shares


Good Category Existing system Revenue-neutral

uniform VAT rate

Zero-rated food 12.1% 13.1%

Standard-rated food, catering and alcohol 12.1% 10.9%

Leisure goods and services (inc. tobacco) 22.3% 22.4%

Domestic energy 5.7% 6.7%

Household goods and services 11.9% 12.7%

Personal goods (inc. clothes) and services 14.5% 14.4%

Private transport 19.2% 18.4%

Other zero-rated goods 2.3% 1.4%


Revenue-neutral move to a uniform VAT rate Effect on welfare


-0.8%

-0.6%

-0.4%

-0.2%

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

1.2%

Poorest 2 3 4 5 6 7 8 9 Richest Average

We

lfa

re g

ain

/lo

ss a

s %

exp

en

dit

ure

Income Decile Group


Distributional impact of changes 2015-2019

Figure 1. Personal tax and benefit measures Figure 2. ‘National Living Wage’

-1,400

-1,200

-1,000

-800

-600

-400

-200

0

200

400

£ p

er

ye

ar

Net income decile group

-1,400

-1,200

-1,000

-800

-600

-400

-200

0

200

400

Net income decile group

Source: Joyce (2015) © Institute for Fiscal Studies


Microsimulation in the UK: TAXBEN

Stuart Adam, IFS

Workshop on microsimulation modelling for fiscal policy analysis

European Commission Joint Research Centre, Seville

23 September 2016

microsimulation in the uk: taxben · tax and benefit systems since 1975 – plus, crucially,...

Documents