Examination of witnesses (Questions 413
- 419)
THURSDAY 16 SEPTEMBER 1999
DR HELEN
BARNES, DR
GILLIAN PAULL
and PROFESSOR IAN
WALKER
Chairman
413. Good morning, ladies and gentlemen. Can
I welcome our friends from the Family Policy Studies Centre. We
have Dr Helen Barnes, Dr Gillian Paull and Professor Ian Walker.
Helen, I wonder if you would not mind saying a few words about
your own work. We have been interested to read the Trial and
error November 1998 piece of work which was very valuable
as background but there have been some changes in some of the
public arguments and debates since and you might like to bring
us up to date. Also could you introduce to the Committee the work
that Dr Paull and Professor Walker have been doing. You have very
helpfully supplied us with a summary CS 43, that is very useful
and perhaps Professor Walker or Dr Paull can talk us through that.
We have not got a lot of time so if you can try and make it as
brief as you can.
(Dr Barnes) I will keep my comments very brief in
that case because I think we have quite a lot of new material
to add to what was presented in the written evidence. I am Helen
Barnes, I am based at the Family Policy Studies Centre. In addition
to Trial and error: a review of UK child support policy,
which was published in November 1998, which I was the lead researcher
on, we have also published a European comparison of child support
systems by Anne Corden based at the University of York. We have
had an ongoing role in the debate about child support reform.
Certainly as part of that we have very much argued for increased
simplicity and transparency and to that extent we are broadly
supportive of the proposals but we have some concerns. One of
our concerns which emerged when looking at the Green Paper some
months ago was the relative equity of the proposals between different
family types and whether, in fact, some families are going to
be made poor at the expense of others. As part of that the Nuffield
Foundation, which funded the earlier review, agreed to fund a
study jointly between the Family Policy Studies Centre and the
Institute for Fiscal Studies, for which Gillian Paull and Ian
Walker work. That really looks at the evidence by modelling from
existing survey data what we know about the incomes of families,
first and second families, models the existing child support regime
and then looks at what the new reforms will mean. Will second
families be pushed into poverty? Will first families be pushed
into poverty to make allowances for second families? Some of the
evidence that we will put forward today looks at that. It also
looks at things like earnings incentives. This is quite a short
self-contained piece of work very much looking at the reforms
themselves and that will be published in the very near future.
Gillian Paull and Ian Walker also have a longer scale study based
at the Institute for Fiscal Studies looking at the long-term behaviourial
effects of the reforms in terms of re-partnering and in terms
of labour market supply. That will start very shortly and it was
also funded by the Nuffield Foundation. That will go on for two
years. I think I have outlined some of our concerns in the written
evidence so I do not need to go over those, you can ask anything
you have in mind. I will hand over to Gillian and Ian.
414. Thank you. Professor Walker?
(Professor Walker) My name is Ian Walker. I work at
the University of Warwick and some of my work is based with the
IFS. I have a history of work on the statistical modelling of
the impact of policy reforms on various aspects of individual
behaviour. My interest in child support goes back some way prior
to the earlier reforms. I have published some work on the impact
of the early reforms on the incentive for individuals to work.
We have come back to the issue, motivated by the new reforms,
and our work is going to be funded by the Nuffield Foundation.
To begin with we began to think about what the characteristics
of the current system were, what we liked and what we did not
like about it, and then began to think about what the impact of
the reforms might be. There is a list of good points and bad points
about the existing proposed systems on the sheet. One of the striking
things about that is that you can marshall logical, plausible
arguments in favour of specific aspects of the current scheme
or the reforms but similarly you can argue against those aspects.
For example, the child support disregard in the child support
system under the new system can be argued as a good thing in that
it might promote compliance but, on the other hand, it may be
argued as a bad thing in that it reduces the incentives for caring
parents to work. There are trade-offs involved in the design of
child support. In order to evaluate what the net effects of the
reforms are you need to do quite a complex statistical analysis
of the behaviour of real people. That is what the Nuffield funded
work is going to be about over the next two years. In the absence
of UK evidence we have been looking at US evidence and there are
a number of striking things about the US analysis. Firstly, separation
basically makes the absent parents better off in terms of their
disposable resources. Typically fathers end up being better off
after a separation whereas mothers end up without the income and
with the children and end up being substantially worse off. In
the US child support has a reasonably large impact on child poverty
amongst the children who end up living with the mother and only
a small adverse effect on child poverty in second families because
the fathers end up being quite rich. The US ought to be quite
informative about the prospects for the current reforms in that
since 1996 the Clinton Welfare Reforms have allowed each state
to go their own way and different states have chosen different
routes for their child support systems. In particular, some states
have dropped the old disregard for child support that was a feature
of AFDC. AFDC was the US equivalent to income support. Some states
have dropped this disregard and some states have kept it. It turns
out that there has not been any difference in changes in compliance
between those two groups of states. If the US evidence can be
translated to the UK context it seems that the disregard for child
support in the welfare system, in particular income support, may
not actually have much impact on compliance. That is a worrying
implication for the reform because one of the things that the
reform will do is that it is going to make caring parents worse
off unless there could be a significant increase in compliance.
So that is the worry we have on the basis of the US evidence.
We have some UK work that has looked at the impact of the reform
on the incentives for individuals to work and there is a table
in our note[3]
which shows you what happens under the reform to the distribution
of working behaviour for a large sample of households taken from
the Family Resources Survey. What happens is that introducing
WFTC, replacing family credit by WFTC, which is about to happen,
is actually good for work incentives, it reduces the proportion
who are not working by about two or three percentage points. The
child support reform itself contributes further to that process,
so that it turns out that the complete disregard for child support
in WFTC under the reform has a beneficial incentive on work which
outweighs the adverse incentive effect of the income support disregards,
so whilst the income support disregard may, on the US evidence,
not promote compliance, at least it does not seem to do too much
damage to the incentive to work. That analysis is somewhat naive.
It takes a bunch of old behavioural estimates, simulates how people
would behave in the labour market and simply runs through a bunch
of computer code to show what the effect on behaviour would be.
It is not in any way based on experimental evidence. So the recommendation
that we have so far is that, actually, we do not know what works.
Not only do we need to know what works, but we actually need to
know how well it works because, as I said, child support design
is about weighing up conflicting forces from particular aspects
of design. In order to find out what works and how well it works,
we need more research. In particular, we need to do more work
on the Child Support Agency's own administrative database which
the DSS has had access to, but nobody else. We need better data
than is currently in the public domain. The Family Resources Survey
is quite extraordinary in that it only tells us about the absent
parents who pay child support and you cannot actually identify
absent parents in that data who do not pay child support; the
survey simply does not allow you to identify that. So the data
that we do have in the public domain is actually censored in a
way which is extremely unhelpful for this kind of analysis, but
we have been doing what we can to overcome that. Finally, in the
process of finding out what works, we need to build evaluation
into the reform and we need to have the reform implemented in
such a way that we can actually deduce from people's behaviour
after the reform is implemented what aspects of the reform are
changing their behaviour. There are a number of things one can
do and, for example, I suggest here that you might, say, randomise
the level of income support disregard to find out what effect
it might have on compliance and work incentives. Gillian is going
to talk about what we have done with the Family Resources Survey.
(Dr Paull) My name is Gillian Paull and I work for
the Institute for Fiscal Studies and I specialise in the areas
of labour market and benefits with particular relevance to family
policy and working mothers with children. In the work we have
done, we used the sample survey that Ian has mentioned, the Family
Resources Survey, and we have overcome the lack of evidence on
absent fathers by combining it with information from another panel
survey where we can actually see couples when they are separated.
In the study we consider all mothers in the UK not living with
the father or some of their children, so this is not just lone
mothers because some of them may have re-partnered, and it is
also not just the CSA caseload because not necessarily all of
these are on income support and some might actually be in working
households. The area we do not cover is caring fathers and absent
mothers and obviously that is something that the Committee might
wish to bear in mind. We consider the impact of the proposed reforms
and WFTC without allowing for any changes in labour supply. That
is, we assume that everyone carries on working exactly as they
are now, so, if you like, you are getting the lower band on possible
changes, assuming no behavioural response. We have been particularly
interested in looking at three things. Firstly, we want to know
how the average weekly child support payment changes. We would
also like to know how the average net income changes both for
the caring mother and the absent father and his family. We have
also looked very carefully at the proportion of families in poverty
and the proportion of children in poverty and how that proportion
changes with the possible reforms. Now, it is very easy to calculate
how the liabilities change. The problem is calculating how payments
change and we do have to say something about compliance. So we
have looked at current compliance in the data and what we find
is that roughly about half of all liabilities are paid and that
roughly about half of absent fathers who owe a positive liability
actually pay something. The way we do this is we look at our results,
and we look at the effect of the reforms at several different
levels of compliance. Firstly, we consider current compliance
levels and nothings changes in the way that absent fathers respond
to their liabilities. We also look at the level of 80 per cent
compliance which is a figure proposed in the White Paper, which
is the Government's target figure, and that 80 per cent, it is
not clear what is meant by 80 per cent compliance. It could mean
that all absent fathers pay 80 per cent of their liability or
it could mean that 80 per cent of absent fathers pay all their
liability and 20 per cent pay nothing. The distinction between
those two is quite important. The latter where we have 80 per
cent paying all of their liability actually has a bigger impact,
whereas if you had everyone paying just 80 per cent of their liability,
it has a smaller impact and that smaller impact lies somewhere
between current compliance and 80 per cent paying everything.
In the figures I will tell you now, I am going to assume that
80 per cent of absent fathers pay everything and 20 per cent pay
nothing. We also look at what would happen if absent fathers paid
their full liability, so that is if we had complete compliance,
and this provides an upper band on the effects. If you look at
the data, we see that of all caring mothers, some 55 per cent
are on income support whereas 23 per cent are eligible for WFTC
and 22 per cent are working at levels which place them above the
level for WFTC. For absent fathers, on the other hand, only 20
per cent are on income support and that is a much lower figure
than that for caring mothers and only 6 per cent are eligible
for WFTC, and that is partly because they have far fewer children
in the household. In fact less than one quarter of absent fathers
have children in their household. Some 74 per cent of absent fathers
are working at levels which make them ineligible for WFTC. The
caring mother households are much more likely to be on income
support than the absent father households which suggests that
there might be the opportunity to reduce poverty in caring mother
households if we can redistribute that income between households.
Turning very briefly to our results, we considered the reforms
in two or three stages. In the first stage we asked what would
happen following the move from family credit to WFTC, that is
to say, we carry on with the current child support system, and
we considered that in isolation. We find that the child support
payment is essentially unchanged from around £40 a week and
the reason for that is that the absent fathers on family credit
have no liability, so the change from family credit to WFTC does
not change their liability. The caring mothers' income increases
marginally from £208 a week to £212 a week with the
change from family credit to WFTC, but poverty among caring mother
families remains at about 38 per cent and about 43 per cent of
children living with just the caring mother who remain in poverty.
In the absent father family, the poverty falls from around 16
per cent to 12 per cent. Poverty in the absent father family is
much lower than in the caring mother family. Suppose we now move
to the White Paper reforms holding compliance at current levels.
The amount of child support paid falls from an average £40
a week to £29 a week and that is quite a sizeable drop. The
caring mother's average income falls by a lower amount, from £212
a week to £209 a week. The reason for that is there is a
benefit cushion and essentially part of the loss of child support
is made up by increases in other benefits. The absent father's
income rises from £377 a week to £392 a week due to
the lower child support liability. There is very slightly lower
poverty for both the absent father and the caring mother's family
but it is a very small change. Overall family poverty falls from
23 per cent to 22 per cent and poverty amongst children, that
is children in the caring mother household and the absent father
household, falls from 37 per cent to 35 per cent. That is the
impact of the White Paper reforms assuming there is no change
in compliance and obviously it is a stated goal of the Government
that it will seek to improve compliance. Let us suppose that compliance
does increase to the 80 per cent target, in that situation the
average child support payment moves back to £41, that is
just one pound higher that it is under the current system at current
compliance. The caring mother's average income moves to £214,
that is only two pounds above the average level of income on the
current system and at current compliance levels. Poverty is changed
very little, it remains at about 21 per cent for the families
and 35 per cent for children. Even if there was full entitlement
there would be very little change in child support pay in net
incomes. We also have a rough estimate of the cost to the Government
of these changes.[4]
Supposing that we continue to move from Family Credit to Working
Families Tax Credit, we estimate that the cost for these groups
of separated families is about a quarter of a million pounds.
That is an annual net change in Government revenues. The cost
of reforming or changing from the current system of child support
to that proposed in the White Paper we estimate to be about £440,000
annual Government revenue if there is no change in compliance
but only about £75,000 if the 80 per cent target level of
compliance is achieved. Finally, I just want to say something
about three specific subjects. I have talked about all separated
mothers at the moment and obviously the CSA client group considers
only caring mothers on income support, that is their main target
group at the moment. We looked separately at what would happen
to the caring mother's income support and we foundagain that their
child support payments fall with the reform, that is from the
current system of WFTC to the reform system of WFTC, from £28
to £20 a week. But if there is improved compliance then the
amount could actually rise to £36 or £39. Improved compliance
can actually have an impact on the net income of caring mothers.
The income rises from £152 to £156 and possibly only
to £160 with improved compliance. Poverty for this group
is quite high, it is currently 61 per cent of caring mothers on
income support on our poverty measure. This will fall to either
55 per cent under the reform or to 52 per cent even if we achieve
the 80 per cent compliance. Similarly, the proportion of children
in poverty is currently 64 per cent and this will fall to 60 per
cent or 58 per cent at best with improved compliance. So the changes
in poverty rates are not very dramatic.
Mr Pond
415. Could you give us that figure again, please?
(Dr Paull) 64 per cent to 60 per cent if there is
no change in compliance and 58 per cent if compliance improved.
A second area we have looked at is the impact of the income support
deduction. A £10 proposed deduction for income support does
actually seem to have an important effect. For the group of caring
mothers on income support the net income without the deduction
will be £5 to £7 lower. Also family poverty will be
eight percentage points higher and the number of children in poverty
will be six percentage points higher without the income support
deduction. The cost to the Government of this deduction we estimate
to between about £150,000 to £200,000 a year. That seems
to be quite a move in poverty compared to the cost.
416. Can I just ask for clarification on that.
Six percentage points or six per cent?
(Dr Paull) Six percentage points. I can give you the
actual figures. On the other hand, for the Working Families Tax
Credit full deduction, we considered the impact of that over the
current £15 deduction in Family Credit. It raises the net
income very marginally for caring mothers but it does not have
any impact on poverty for the simple reason that mothers who receive
the Working Families Tax Credit are not really in poverty. Finally,
we also looked at the impact of including step children in the
absent father's allowance in assessing his liability. We estimate
that the exclusion of step children from this calculation would
raise child support payments by around £6 to £9 a week.
It would raise net income by about £4 for the caring mothers
of absent fathers with step children. It would fall by roughly
the same amount for the absent father's family with no actual
overall change in poverty. The inclusion of step children causes
a slight switch in resources from the absent father to the caring
mother. Our overall conclusion is that the reforms have relatively
little impact even if compliance changes dramatically, which might
be for better or for worse, but it is not likely that you are
going to achieve a great reduction in poverty. On the other hand,
you are not going to be plunging large numbers of absent fathers
or caring mothers into poverty. It is true that the child support
liabilities do fall but if compliance is increased then the actual
payments will stay around the same amount. There is very little
impact on net income and the poverty levels regardless of compliance.
This is partly due to the benefit cushion, that benefit payments
change in allowance for the change in child support payments.
It is also due to the fact that where payments change they are
not sufficiently substantial to move families with children out
of poverty or into poverty. More importantly what this work stresses
is the general issue of whether absent fathers are actually sufficiently
well off to be able to move the corresponding caring mothers in
or out of poverty to a great degree. It also suggests that if
the goal is to reduce poverty amongst caring mothers with children
our best hope might be to look towards the effect of changes in
incentives to work behaviour and not to changes in the child support
payments themselves.
Chairman
417. Thank you. You will understand that is
quite a lot for us to take in.
(Dr Paull) I do. I have some tables which I do not
know whether it would be useful for me to photocopy and give to
the Committee.[5]
418. It would be very helpful if we could have
a paper for the reason that it is easier to put a wet towel around
your head and make sure you understand it all. That sounds a very
valuable piece of work, it is just very difficult to know where
to start constructively criticising it just listening to it for
the first time. Is it possible to have copies of the key figures
and the key conclusions?
(Dr Paull) Yes, I can give you a copy of everything
I have just said.
419. Do I understand that this is really fresh
evidence and work that has just been done?
(Dr Paull) Yes, literally in the last few days.
3 See Ev. p. 167. Back
4
Subsequent to the hearing, the cost estimates have been substantially
revised and are presented in supplementary material. See Ev. p.
177. Back
5
See Ev. pp. 176-180. Back
|
