The DWP on welfare reform: good on the generalities, weak on the details

“The welfare state is now a vast, sprawling bureaucracy that can act to entrench, rather than solve, the problems of poverty and social exclusion.”

The above quote does not come from a think tank publication, but from a policy paper by the Department for Work and Pensions (DWP). Compared to the pamphlets the DWP released under the previous administration, this one is a delight. It describes the failings of the present welfare system – mind-boggling complexity, multiple benefit types working against each other, steep withdrawal rates – with a boldness one would never expect to come from within the political machinery. It also outlines an alternative strategy: merge most non-contributory benefits and tax credits into a single “Universal Credit” (UC) with a uniform taper rate.

But as good as the paper is on the generalities, it stops at the sound bite level when it comes to the more unpleasant trade-offs in welfare design. The paper’s basic dilemma is apparent in this quote:

“Reforms could:

  • improve work incentives by reforming the way in which benefits are tapered as incomes rise and allow people to keep more of their earnings;

  • be fair and targeted to those most in need through tapers which focus payments on those on the lowest incomes, while maintaining levels of support for those out of work.”

Both of these points make sense in their own right. But in combination, they blatantly contradict each other. For any given level of income replacement, lowering taper rates improves work incentives, but it also makes benefits less targeted by spreading entitlement to people further up the income distribution. Raising the taper rate achieves the opposite.

Let’s put some numbers on this. Suppose a workless single household receives £280 per month in Income Support, and, just to obtain a round number, £420 in Housing Benefit. (£420 is less than what can be claimed in middle-sized cities like Bath, Birmingham, Canterbury or York.) Let’s ignore Council Tax Benefit and any possible add-ons. The DWP-paper emphasises that there is no intention to cut out-of-work support, so this household would be entitled to £700 in UC.

The rest depends on the UC taper rate. If set at 75%, our household would have to earn £933 per month to come off benefits. This corresponds to an equivalised income of £1530 (=£933/0.61), which would place the household somewhere close to the 40th percentile of the income distribution. UC would then be fairly targeted, but an implicit marginal tax rate of 75% is still a heavy penalty on work.

Lowering the taper to 55% would improve incentives, but it would turn UC into a ‘middle-class benefit’. Our example household would now have to climb up to about the 60th percentile of the distribution to come off UC. Working from the reverse end, it would be possible to calibrate UC in such a way that it trails off as the household reaches the 25th percentile of the distribution. But this would require a taper of 95%, a rate at which only devout Calvinists would care to advance in the labour market.

UC would be a huge improvement over the present system because of its greater clarity and predictability. But it would not solve the basic dilemma that within the present system, we can only choose between stifling work incentives, and pouring out huge sums of money to well-to-do people who do not need it. When it comes to welfare reform, simplification can only be a start.

19 thoughts on “The DWP on welfare reform: good on the generalities, weak on the details”

  1. Posted 10/09/2010 at 12:48 | Permalink

    You are overlooking the Laffer Curve, which applies to means testing in exactly the same way as it applies to taxation. If the top of the Laffer curve for employment income is (say) 60% (and I believe it is) and Employer’s NIC is about 11%, then mathematically, the cost-minimising marginal withdrawal rate 50%, which can easily be dealt with via the PAYE system using K-codes.

    Further, the DWP & the Tax Credits system spend vast amounts of money very inefficiently, it would COST LESS to pay out Universal Benefits to EVERYBODY at approx. Income Support rates – that’s simple maths. Chuck in a much lower marginal rate and you are bound to come out with a lower overall cost.

  2. Posted 10/09/2010 at 12:54 | Permalink

    Mark – what would you do about Housing Benefit and subsidised rents (i.e. social housing)?

  3. Posted 10/09/2010 at 13:16 | Permalink

    Exactly. The problem isn’t Income Support. If that was the only benefit, it could be tapered at a rate much smaller than 50%, and it would still not become a middle-class benefit.

  4. Posted 10/09/2010 at 14:03 | Permalink

    The cost of housing is far from uniform even within a single city.

    Thus, before folding in HB, one must consider localising the setting of the UC rate just as one should end national pay scales.

    Further, why should the taper itself be uniform?

  5. Posted 10/09/2010 at 14:29 | Permalink

    On the issue of localising, the paper is just as contradictory. It first says:
    “The UK has a highly centralised system of benefit design and delivery. […] a number of other countries […] operate more devolved welfare systems, which can stimulate innovation and ensure that systems are more aligned to local circumstances.”
    And then goes on to say:
    “The Government will continue to work closely with the devolved administration in Northern Ireland to seek to maintain a single system across the United Kingdom.”

    Remarkably, these two statements are from two consecutive pages.

  6. Posted 10/09/2010 at 18:48 | Permalink

    RW, housing benefit is easy!

    One third is paid to ‘private’ landlords. it is far cheaper to build a bit more social housing (which in cash terms costs nothing, and worst case, the council can flog it off at a profit later on), but failing that, they can just vastly reduce the cap.

    2/3 is paid by DWP to local councils (or housing associations, which are also ultimately state-owned), so IT IS MADE UP FIGURES. if social rents were set to £nil, then that 2/3 of the DWP’s HB bill would sink to zero! As it happens, the income from social housing just about covers the cash costs anyway (and rents could be collected via PAYE as well, so would adjust down if people’s incomes fell).

  7. Posted 10/09/2010 at 19:52 | Permalink

    “Remarkably, these two statements are from two consecutive pages.”

    But who says the Government wants to “stimulate innovation and ensure that systems are more aligned to local circumstances”?

    It appears they do not.

  8. Posted 11/09/2010 at 11:50 | Permalink


    Adjusted down if earnings fall, but I think it is important to consider adjusting up until market rates are applied so that either tenants either move out OR the rental income can cover HB in the private sector for those who would otherwise be so housed*. However it does imply two mechanisms – one for those earning, another for those not and a mix for many unless you introduce negative Income Tax.

    * Just because the State could cover costs of housing does not mean the State should become an even bigger Landlord. Mobility is already impacted negatively as it is by the distortions created.

  9. Posted 11/09/2010 at 16:21 | Permalink

    Mark, interesting idea on HB. However, central government has been using local authority rent surpluses to subsidise HB for the last 15 years. Likewise housing associations use HB as a de facto subsidy – just see all the squealing about the relatively minor restrictions to be HB announced in June.

    Also whilst no government would allow a large HA to go under, it is central to the funding of social housing that HAs appear to be private: in this way their borrowings don;t count as public funding.

    Its therefore very complex web and simple ideas – whatever their virtue – have lots of consequences.

  10. Posted 11/09/2010 at 16:23 | Permalink

    Tim, you might be right. One of the big debates in Conservative circles is whether the Coalition really mean it when they promise fundamental reform. Many people remember what Blair and Brown – and frank Field – said in 1997-8.

  11. Posted 11/09/2010 at 20:23 | Permalink

    TCLP: [Social rents would be] adjusted down if earnings fall, but I think it is important to consider adjusting up until market rates are applied so that either tenants either move out…”

    Yes of course! One of the greatest ideas I ever nicked off somebody else was to give social tenants a K-code for PAYE so that they pay an extra 20% of their wages in “rent”. If they earn little, they pay little; if they earn a lot, they will find it cheaper to rent privately (or buy their own home). Or maybe they’ll bank on safety and happily overpay a bit in the good years in case they lose their jobs again. It’s all good.

    PS, this is not a new idea, other countries do it.

  12. Posted 11/09/2010 at 20:28 | Permalink

    TCLP: “* Just because the State could cover costs of housing does not mean the State should become an even bigger Landlord. Mobility is already impacted negatively as it is by the distortions created.”

    As a free market liberal, I don’t see a problem. Let’s imagine the state built ten million new units of social housing where most needed. Anybody would be able to live anywhere in the country he wanted. There’d be social mobility, labour mobility, an end to the house price bubbles etc etc and rents overall would drop, so the local councils could charge a market rent which is still affordable for most etc etc. The rents could be used to reduce income tax or VAT, etc.

  13. Posted 11/09/2010 at 20:33 | Permalink

    Peter, I am never happier than when debating this sort of thing, but first let’s agree on actual hard facts.

    Click my name for my own ‘crash course in social housing’ which hopefully sets out the background.

  14. Posted 12/09/2010 at 16:55 | Permalink

    Mark, thanks for the link. I would be happy to correct the factual errors or send you a reading list.

  15. Posted 13/09/2010 at 10:58 | Permalink


    Re: 10m units. How would one administer the allocation of such? Define “where most needed”. Do you trust the bodies to know what that is? How long before “where most needed” changes and will the bodies know or care? What of land acquisition and contract allocation?

    If you see subsidised rents, how can we then determine market rates for rentals with a 10m unit gorilla sprawled across the nation?

  16. Posted 13/09/2010 at 11:02 | Permalink

    Tim Carpenter makes some excellent points. Another problem is that social housing is means tested, by a points system etc., so acts as a disincentive to work even before the housing benefit gets paid.

  17. Posted 13/09/2010 at 13:11 | Permalink

    Tim indeed makes an excellent point. Didn’t have Hayek say something about the fallacies of central planning?

    The manner in which social housing is allocated is indeed issue, and one that the government seem to be aware of. However, it too is a very complex issue that goes to the very heart of the purpose of welfare provision.

  18. Posted 27/09/2010 at 15:20 | Permalink

    Tim, how about applying commonsense to each of your rhetorical questions?

    Q “How would one administer the allocation of such?”
    A – By price and/or waiting list.

    Q “Define “where most needed”.”
    A – Look at market signals – where prices high = most needed.

    Q “If you see subsidised rents, how can we then determine market rates for rentals with a 10m unit gorilla sprawled across the nation?”
    A – Who said ’subsidised rents? Just let them out for ‘as much as the local council can get’, which by definition is ‘market rent’.

    And so on.

  19. Posted 01/10/2010 at 13:07 | Permalink

    Having been stuck on benefits for many years now, for me, and many others, the fundamental problem is housing.

    Having been self-employed and in receipt of HB and WTC, I think that any housing element of a Universal Credit should be available to use against a mortgage as well as rents – but yes I understand the localisation issue here.

    It’s difficult saving on a low income, but I believe a NON-MEANS TESTED saving option should be permitted so that you could save for a mortgage deposit for example. Currently housing benefit reductions disincentivise savings.

    I think schemes to allow the low-income to build there own homes should also be introduced – is it such a crazy idea?

Comments are closed.