Universal Benefit - replacing JSA, IB, ESA, and IS
Presenter: Lawrence Kay
Recognised as a major influence on the thinking of a future Conservative government, the Policy Exchange is moving beyond Freud to look at one of the enduring problems of the welfare state: the ever-increasing complexity of benefit rules, and how they might be simplified. As Theresa May acknowledged in her own appearance at the conference, research on this is in the earliest stages, and the session was mostly a debate rather than a presentation of results. Taking the idea of a single, universal benefit as a starting point, the discussion quickly revealed why so many ministers have come in intending to simplify the benefits system and left with something even more complex in its place.
How complex is the benefits system?
Given that fully understanding and offering advice on the benefits system is a full-time job, it's pretty darn complex. Lawrence used some rough proxies, including:
- The Child Poverty Action Group guide to benefits and tax credits was 400 pages long in the early 90s, and is 1500 pages long now
- The Public Accounts Committee found 30 linkages between different benefits in a study some years ago
- The National Audit Office looked at complexity and internal DWP guidance – 14 volumes of guidance
- The NAO also found that the past 10-15 years had seen significant changes to benefits in every single year
- The Decision Maker's Guide, which acts as the Bible for DWP benefits, runs to 15 volumes
How did it get so complex?
Lawrence identified 'fairness' as the main reason for increasing complexity; adding wrinkles and exceptions so that the system adheres to people's sense of justice by giving help to those who deserve it and protecting state finances from those who don't. Responding to specific circumstances (e.g. seasonal workers in Scottish mills) often necessitates flexing of the system.
Although unstated, there's an obvious comparison with the legal system to be made – the UK common law system is hugely complex, and tends to get more so over time.
Why simplify?
Lawrence's primary argument for simplification was that the complexity of the current system reduces the willingness of benefit claimants choosing to move into employment, thus making the job of JCP advisers and providers much harder. In the absence of quantitative research on precisely how much it reduces job entry, Lawrence used complexity in the tax system as a comparison point. Various US research projects were deployed to show that people were far more likely to make bad decisions in a more complex system.
How simplify?
The starting point for simplification was Roy Sainsbury's proposal of a single universal benefit to replace JSA, ESA, IS and IB, paying a flat rate of £60-70 / week. Tax credits and other benefits such as DLA, SDA and so on would be unaffected. Housing Benefit was also not covered. In universal benefit, all JCP adviser interviews would be work-focused.
Discussion - I do not think it means what you think it means
Much of the discussion was devoted to the wholesale demolition of the universal benefit proposal. Objections included:
- Repeating the past: what would stop the new benefit from increasing in complexity over time, in exactly the same way that the old benefits did?
Lawrence's answer to this was that government's desire to maintain simplicity would suffice for 15 years or so. He also pointed to the introduction of universal benefit in New Zealand in 2010 as an example of the achievability of simplification - The complexity is there for a reason: it matches the complex needs of the people being helped by the system
Giving front line advisers the ability to target funding and support to people on an individual basis would allow them to meet these needs. This flexibility would also remove the rules that prevented advisers from helping customers - The proposed flexibilities could place a huge burden on advisers, who would need substantial training and support to take advantage of them
- Flexibilities could also undermine the conditionality (i.e. the fortnightly signing, job search requirements etc.) that underpinned much of the success in reducing unemployment prior to the current recession
Lawrence clarified that adviser flexibilities would still need to operate within a solid framework of rights and responsibilities - The universal benefit would not replace most of the complexity in the current system, as DLA, industrial injuries, Housing Benefit, capital rules, eligibility requirements etc. would all remain in place
- Even if the entire benefits system was successfully simplified, the complexity of government engagement with claimants and employers would remain, because of the fragmented, silo nature of government delivery through many different departments
- Replacing benefit rules with adviser flexibility would result in inconsistent and often unfair support for customers with particular needs, dependent on their adviser's current mood, availability of funds, caseload pressures, local practices, and simply whether they make a good impression or not. Additionally, giving that level of power to advisers may damage the relationship and trust between advisers and claimants, and potentially result in abuse of power.
Estimates from a few years ago are that 20% of benefit decisions are wrong. 80% of eligible tax credit recipients don't claim them. The mere existence of full time benefit advisers proves that the system only works properly when it's pushed on by people who know what they're doing. Given this, the system is already inequitable in practice.
A number of participants pointed to the long history of failed attempts to simplify the benefits system, from Age Concern's work in the 1970s onwards. The consensus view seemed to be 'don't start a land war in Asia'.
It wasn't all demolition though. There were also some suggestions for improvements or alternatives:
- Moving payment of Housing Benefit into the DWP's remit
It was pointed out that HB was originally a central government responsibility, before being devolved to Local Authorities - Look at the complexity of transitions between benefits and into work, which is what really stops people from moving off benefits
- The Frank Field suggestion of making disability and health support benefits independent of employment status and work readiness. This, coupled with universal benefit, ensures that those with specific health needs still receive extra support, but removes the disincentive to entering work
- Housing Benefit is only paid for four weeks after job entry. Making work appear worthwhile for the long-term unemployed could necessitate paying HB for at least a year
- Transition into work can take 1-2 years for the long-term unemployed. Changing benefit rules and programmes to allow them to undertake training without hitting obstacles such as the 16 hour barrier would help enormously
- A one-button approach to job entry, with a person's benefits being suspended rather than stopped for their first month in a job, and instantly reinstated without having to go through the process of reclaiming should the person leave employment again
- Make it easier for employers to hire people in the first place – there's apparently research showing that very small employers would be far more likely to hire people and grow their businesses if the employment system wasn't so daunting to navigate from their own side
Political bonus question – in response to a final question about the Conservative welfare agenda, the suggestion was made that, owing to the recession and the huge government deficit that resulted from it, the thrust of Conservative policy in a future government would be cost-cutting. Canada was mentioned as an example of the likely mood.
- 4058 reads





Comments
As an aside, it seems surprising that there isn't a way of deriving a single figure for a system's complexity that could be applied to various systems of rules such as legal systems, benefit systems, process manuals etc. Using this, a baseline could be established of how much someone could be expected to understand in a given time, all other things being equal. It would be even more interesting if this could be derived for different skill levels, rather than just the 'average' person. It may be that contextual understanding makes this impossible, but it would make it much easier to set targets for and measure simplification in a meaningful way.
Quality training is the key.
I have been giving legal advice on welfare benefits since the early 1990's and agree that the system has become more complex.
Simplification has been a goal of successive governments but changes, whether piecemeal or wholesale, have tended to make things more difficult for advisers.
An example is the recent introduction of employment and support allowance meaning that 2 different systems for sickness benefits will be running at the same time for the next 4-5 years.
Even so, simplified structures for benefit can be still be learnt which advisers can use to give informed advice.
Neil Arnott
socialwelfaretraining.co.uk
Daniel, complexity is...complicated. Even plain old entropy is murkey.
http://www.cscs.umich.edu/~crshalizi/notebooks/complexity-measures.html
A nicely intractable example of complexity in the benefits system is the tax credits circularity:
From this it's quite easy to invent cases where couples are paid say, which entitles them to maximum tax credits, which together with their other income disentitles them to income support, reducing their tax credits so they are entitled to income support and so on. Such cases will usually be couples with a partner working 16+ hours and a claimant working less than 16 hours (or aged 60+ and claiming pension credit). But there are other possibilities.
Neil, if only it were just two systems. Simplified and off the top of my head:
Some of these are taxable (and count as income against tax credits), some not. This can have implications for better-off calculations years into employment.
As regards 'fairness' as an excuse for complexity, many of the complications created by recent benefits changes have resulted in grossly unequal entitlements for people in very similar situations.
Any universal benefit will have to play nicely with the tax credits system if claimants and ex-claimants are to receive fair treatment (again using fair in the very limited sense of similar situations producing similar outcomes).
I fail to understand why tax credits have to go through HMRC in the first place. So what if they're income-related? HMRC can still share the necessary data with the DWP. Sticking all the benefits through a single point of contact would at least be a start.
Re: complexity theory. I took a course on this as part of my Maths degree ages back, although it's primarily a Computer Science subject. Algorithmic complexity tends to look at theoretical computability rather than human processing capability, so it's even less applicable than it may seem from the article you linked. A possible way forward might be representing the benefits system as a decision tree, and calculating the number of decision points and the number of possible end states. Relating this to human capability would involve creating lots of rule systems with known decision trees and end states and getting test subjects to try and use them under experimental conditions. Playing around with different presentations of the rule system would also show how the presentation of rules (i.e. the context) affected people's performance. I still think it should be possible, but there may be practical issues around the same system having multiple representations with differing levels of apparent complexity.
I just had a really amusing idea on this, with regard to the single-point-of-contact problem. The issue is not that benefits are administered by different departments. The issue is that the front-end, the interface with the user, is not co-ordinated.
So: create a secure standard for electronically communicating benefits claims and responses across all government departments, local authorities etc. Do this instead of creating a claiming front-end for each benefit on the web, or in a council office or wherever. Once that's done, a universal benefits claim front-end is far, far simpler. As well as calculating likely benefit entitlements, benefit calculator software would actually be able to submit user's claims, as would benefit advisers in CABs, JCP offices, council offices and so on. Submission and certification of evidence to entitlement would be interesting but not impossible.
As an additional bonus, setting this up would probably save government departments a substantial amount of money in creating and administering their own claiming front-ends. Of course, it depends on the achievability of creating a schema that includes all the information required to process any kind of benefit claim. However, this isn't so difficult - UBL is an example of an XML Schema that includes the majority of commercial transactions.
No, this wouldn't solve the actual benefit complexity problems. Yes, it could potentially blow up spectacularly if faced with tax credits circularity. But then, so would most people.
try treating it as a fractal problem, you may be able to calculate the fractal dimension of the system.
Tax credits were put through HMRC for political reasons, to disguise their benefits nature. It was not a good idea (though, strangely enough, DWP did become a bit more friendly to 3rd party organizations once they realized their powerbase was under threat).
Governmentwise, there is quite a lot of work being done on joining-up things at the administrative level - it's at the logico-legal level that things don't mesh, fundamentally because benefits entitlements tend to be day-to-day whereas tax credits are calculated using income (historic/current/anticipated) over pairs of tax years - and because tax credits law is framed in the context of tax law, rather than social security law.
I'm not sure that your proposals would actually generate figures more useful than those you've given already (book pages etc), though the exercise would very likely throw up a fair few useful insights along the way - especially if it fed into something that could help measure the costs of change. 'Representing the benefits system as a decision tree' is pretty much my job description, btw.
I thought the fractal dimension suggestion was a particularly obscure combination of maths and benefit system gallows humour, but real world data can apparently have fractal dimension applied to it. I still think someone's having a laugh though.
Tim, thanks for the info - I'm a bit vague on the benefits system, so it's very useful to be told what's what. The idea on analysing complexity would only be useful in the context of measuring the potential impact of simplification - by comparing the decision tree for a proposed new system it would be possible to see if things were going to be improved enough to make it worth going for or not.
Lawrence is on holiday at the moment, but I might try and put him in touch when he gets back, if you're interested.
I've been pointed to an IFS report that examines how difficult it is to figure out how much time, effort and money it takes for claimants to make their claims and comply with the claim requirements. Full details here
The Centre for Policy Studies has just released a report on benefit simplification, arguing for a much more radical approach: putting all government benefits under the control of one agency, with claims going through a single website. The idea is that this will allow the benefits to be administered more efficiently and brought into line with each other, so that incentive structures for people to get back into work start working properly. Here's the press release.
I'm closing down this discussion. Please comment on the Centre for Social Justice proposals if you'd like to add to the debate.