Letter to the Clerk of the Committee from
Professor Ian Walker (CS 48)
1. At the Social Security Select Committee
session on 14 September I promised to deliver some estimates of
the impact of the child support disregard in the Income Support
(IS) system.
2. In particular the Committee expressed
an interest in knowing the effect of the size of the IS disregard
on work incentives. We could generate results on any aspect of
behaviour for more or less any scenario required but I present
below what seem like good ones to focus minds.
3. The results for caring mothers are as
follows (and supersede the ones reported in the notes CS 43 that
were delivered on the day):
| Per cent | Per cent
| Per cent | | | Child poverty |
| NW | PT | FT
| E(h) | Net income (£)
| per cent |
| Compliance = 40 per cent |
| Pre-reform (no IS disregard) |
| 47.9 | 26.0 | 26.2
| 13.8 | 216.3 | 22.7
|
| Post reform with £10 disregard
|
| 45.8 | 31.3 | 22.4
| 13.4 | 217.4 | 22.2
|
| Post reform with £15 disregard
|
| 46.4 | 31.3 | 22.4
| 13.3 | 217.3 | 22.1
|
| Compliance = 80 per cent |
| Pre-reform (no IS disregard) |
| 45.1 | 30 | 25
| 14 | 228.5 | 17.9
|
| Post reform with £10 disregard
|
| 42.2 | 37.3 | 20.5
| 13.6 | 230.4 | 17.9
|
| Post reform with £15 disregard
|
| 43.6 | 36.1 | 20.3
| 13.3 | 229.8 | 17.4
|
4. NW = not working, PT = part time work, FT = full time
work, E(h) is the expected hours of work per week (which you could
think of a summary measure of labour supply and is basically the
weighted average of PT and FT). Net income is the caring parents
net income including CS and is "equivalised" using the
HBAI method where a couple=1. The child poverty rate is the rate
amongst "first" families including those repartnered
mothers using the 1/2mean income (not allowing any deduction for
housing costs) rule.
5. Pre reform assumes that WFTC has happened and the
WFTC take-up rate is 90 per cent. In each case the post reform
scenarios assume that WFTC has happened, take up is 90 per cent,
the WFTC disregard is infinite (ie all CS is ignored), and the
Housing Benefit CS disregard is also infinite (the DSS/CSA evidence
suggested that this would remain at £15 but our information
is that this is likely to be set the same as WFTC—ie a full
CS disregard).
6. The interpretation is as follows:
— The reform with £10 IS disregard (holding
compliance at 40 per cent) increases labour market participation
but decreases FT work and the net effect is that overall labour
supply would fall (from 13.8 to 13.4), net income rises a little
and poverty rate falls a little.
— The £15 disregard tends to reduce participation
a little relative to the £10 case, and overall labour supply
falls further to 13.3.
But if compliance rises the effects would look different:
— at 80 per cent compliance the £10 case would
have a labour market participation rate of 57.8 (compared to 52.2
per cent) so about 12.5 per cent of non-workers now work—this
is close to 100,000 mothers!
— The higher disregard mutes this effect a little—the
labour market participation rate is now 56.4 per cent—so
about 8 per cent of nonworkers will want to work—say about
60,000. The poverty rate is a little lower because the extra £5
is enough to lift a few more mothers and their children (about
another 10,000) out of poverty. But life is full of tradeoffs!
7. Pre reform assumes that WFTC has happened and the
WFTC take-up rate is 90 per cent. In each case the post reform
scenarios assume that WFTC has happened, take up is 90 per cent,
the WFTC disregard is infinite (ie all CS is ignored), and the
Housing Benefit CS disregard is also infinite (the DSS/CSA evidence
suggested that this would remain at £15 but our information
is that this is likely to be set the same as WFTC—ie a full
CS disregard). You should let me know if the Committee would like
anything further. The work is being written up right now and will
appear in Fiscal Studies in December—we aim to present
it to a seminar, where DSS people will be present, on Thursday
30 September.
8. Finally, I understand that DSS/CSA are commissioning
data collection and research analysis in the near future so maybe
we will learn "what works" and how well in the not too
distant future.
23 September 1999
| 
