Giovanni Tonutti, Policy in Practice, presented his work on Using benefits data to assess the impact of welfare reform in London at the International Conference for Administrative Data Research, Queens University, Belfast on Friday 22 June 2018.
This conference is aimed at researchers who use administrative data to better understand populations and societies. His presentation falls under the conference theme of The World of Work which focusses on the labour market experience of those in, and out, of work.
For more information visit www.policyinpractice.co.uk, email hello@policyinpractice.co.uk or call 0330 088 9242.
Using benefits data to assess the impact of welfare reform in London
1. Policy in Practice Using benefits data to
assess the impact of
welfare reform in London
Giovanni Tonutti
2. Outline
1. Introduction to Policy in Practice
2. Welfare reform in London
3. Findings
a) Benefit Cap: what effects on employment outcomes?
b) Self-employment and Universal Credit
4. Conclusion
3. We make the welfare system
simple to understand, so that
people can make the decisions
that are right for them
We help people toward
independence by making
the welfare system simple to
understand.
6. Welfare Reform
• Ongoing and unprecedented reform
process led by the DWP during the last 3
parliamentary terms.
• Objectives:
• Simplify the benefit system
(Universal Credit)
• Reduce costs (Benefit Cap)
• Improve work incentives
(Universal Credit, Benefit Cap)
7. Gap in the literature
Most analysis on welfare reform is static (survey data, typically FRS)
and backward-looking.
Our research questions:
1. What’s the impact of the benefit cap on employment outcomes?
[behavioural responses]
2. How is Universal Credit likely to affect self-employed households
on low-income? [forecasting]
9. 8,828 households affected in July 2017
• A cap on amount of benefits a family
can receive (from November 2016,
£23,000 a year for families with
children in London)
• Exemption granted for those
households working a certain threshold
of hours.
• Average reduction in their housing
benefit is £60.07 per week
• 18,362 children affected
• 61.7% are single parents
12. Has the benefit cap improved the
employment outcomes of affected
households?
13. Two sample groups
Sampling technique:
1. Treatment group: all households with income from benefit above the
£23,000 threshold as of April 2016.
2. Control group: all households with income from benefit close to the benefit
cap threshold.
• Exempted households excluded from the cohort
Analysis:
• DIF in DIF of the employment outcomes across the two groups in the
months before and after the lowering of the cap
14. Higher movement into work
• Households subject to the
benefit cap showed a 3.5 pp
higher change in employment
rates than families in the
control group
• The analysis fails to control
for the additional support from
local authorities received by
households capped
17. 10% of working age households on low income are self-
employed
• Outer London boroughs
see highest rate of low-
income self-employment
• 54.9 % of self-employed
households are private
renters
• Average monthly
earnings £672.47 vs
£840.79 of other
households in
employment
20. Self-employment & UC: Minimum Income Floor
• An assumed level of earnings for self-
employed people in establishing
benefit entitlements
• If actual self-employed earnings fall
below this, UC calculated assuming
this amount is being earned
• Applies after 12 months of self-
employment
Borough UC roll out
Southwark Applicable
H&F Applicable
Enfield Applicable
Tower Hamlets Applicable
Sutton Applicable
Croydon Applicable
Lambeth Dec-17
B&D Mar-18
Ealing Mar-18
Waltham Forest May-18
Barnet May-18
Islington Jun-18
Harrow Jul-18
Haringey Oct-18
Hackney Oct-18
Greenwich Oct-18
Brent Nov-18
K&C Dec-18
Camden Dec-18
21. Average self-employed household faces a £844.67/month
gap between earnings and MIF
Over 90% face a shortfall
On average, an extra 26 hours/week at NMW
needed to overcome this shortfall.
78.2% of those
facing a shortfall
have been self-
employed for 12
months. MIF will
apply immediately
22. £344 p/m worse off under UC
£2,233
• Self-employed among
the worst affected
group as UC rollout
• Awareness among
claimants?
Research to focus on choices self-employed households on low-income make. Are
they entrepreneurs or adapting to new labour structures (i.e gig economy,
freelancing)?
23. Conclusions
• Positive impact of the benefit cap on employment outcomes. Also as result
of LAs’ support to residents affected. On the other hand, households for
whom employment is not an option faced a worsening in living standards.
• Self-employment as a popular option for low-income families in London.
Increase in income not keeping up with increase in living costs and rent.
With UC, the support available to this group will be significantly reduced.
• Admin data holds a huge potential in driving operational decisions,
particularly relevant for local welfare providers (local authorities, charities
and other third sector organisations) .
24. Next steps..
• Final phase of findings in July 2018
• An interactive public dashboard
• Project has secured funding for an
additional 18 months
• Scrapping the surface
• Homelessness key issue in London
– can predictive analytics be the
key?
So today’s agenda
Why you are here
The difference between the two sessions today
What we need from you and what you get in return
Systematic, Scalable, Causal analysis
One big data store – Query across London
Different policy scenarios – Universal Credit and self-employment
Benchmarking – See what’s working, and where the impacts are
Longitudinal - Tie Cause and effect
TA / EA by borough.
Next Steps:
DWP Letter
Your participation in the project
Universal Credit data
Intros :Go round the room, introduce yourself and say what you want to get out of today, what would you like to see the analysis or software do? I have a full presentation but can tailor it to your needs.
Policy : l’ll quickly recap the welfare reform changes that are still to come in this parliament, that we know of, and share some of the analysis we’ve done on the national impact that welfare reforms have had to date.
Local picture: More interestingly, I’ll then focus on the local picture of welfare reforms and share some examples of work we’ve done with other clients. I’ll demonstrate why every household matters by showing you the different impacts of reforms on next door neighbours.
Our approach: By now you’ll want to know how we do the analysis so I’ll take you through our approach and show you a dataset
CTRS: I have some examples of the CTRS modelling work we’ve done for other clients to show you
Software: And then we’ll take a look at our software
Feedback from other clients and frontline advisors
Next steps
We believe people and organisations make better decisions when the welfare system is simple to understand, and they can track the effectiveness of policy.
Our data models, policy algorithms and access to administrative data allow us to combine national and local policies across four government departments, and track the impact of policy on hundreds of thousands of households over time.Advisors use our intuitive software to empower their residents to make decisions related to work and budgeting that are right for them.Local organisations use our forward-looking analytical services to pinpoint support, act proactively and track the impact of interventions on reducing future demand.
October @ £26k
737 = Children fell into poverty?
How many self-employed are there?
Self-employed households: Average monthly reported earnings went up slightly, from £641.93 to £669.72 (+4.3%)
Employed households: Average monthly reported earnings went more significantly, from £762.52 to £822.52 (+7.9%)
Juan – can you do this xyz
Average shortfall between MIF and self-employed earnings is £844.67, or 26 hours/week at the NLW (£7.50). Note this figure excludes B&D
Of the 41,194 households (excl. B&D) that are self-employed as of July 2017, just 3,670 earned on or above their applicable MIF. The remaining 37,524 would receive benefits based on the MIF notional income that applies.