icm2re logo. icm2:re (I Changed My Mind Reviewing Everything) is an 

ongoing web column  by Brunella Longo

This column deals with some aspects of change management processes experienced almost in any industry impacted by the digital revolution: how to select, create, gather, manage, interpret, share data and information either because of internal and usually incremental scope - such learning, educational and re-engineering processes - or because of external forces, like mergers and acquisitions, restructuring goals, new regulations or disruptive technologies.

The title - I Changed My Mind Reviewing Everything - is a tribute to authors and scientists from different disciplinary fields that have illuminated my understanding of intentional change and decision making processes during the last thirty years, explaining how we think - or how we think about the way we think. The logo is a bit of a divertissement, from the latin divertere that means turn in separate ways.


Chronological Index | Subject Index

Notes on universal discredit

Isabelle’s journey through the roll-out of UC for the self employed

How to cite this article?
Longo, Brunella (2018). Notes on universal discredit. Isabelle’s journey through the roll-out of UC for the self employed. icm2re [I Changed my Mind Reviewing Everything ISSN 2059-688X (Print)], 7.11 (November).

How to cite this article?
Longo, Brunella (2018). Notes on universal discredit. Isabelle’s journey through the roll-out of UC for the self employed. icm2re [I Changed my Mind Reviewing Everything ISSN 2059-688X (Print)], 7.11 (November).

Preamble

The subtitle of the first version of this article was: About making judgments the Cambridge way. Then I realised the regular reader of this column would have thought I am talking (again) about miscarriages of justice, no fault evictions libelled by senior civil servants and used to train the Judiciary, and other dickensian misadventures.

Not at all. So I felt I'd better change the subtitle and be more precise: the article is reflecting on what is going on in the automation of welfare benefits services in the United Kingdom, a massive socio-technical experiment impacting the life of millions of people.

The article's focus is on the very basics of information design and engineering: what can and should be considered, learned, taken into account and by whom when we want to make change happen on a large scale leveraging on technologies of information and communication.

That's not cricket

Not long ago I couldn't help overhearing a pub conversation about what would be to employ a proof tool, called Isabelle, to fix some English cricket recurrent problems.

Ah, ah! My first reaction was to join the conversation and ask who would be the cricket champion or team needed to carve the tool.

With the 1996 words of Isabelle's leading author, Larry Paulson, "proof tools must be well designed if they are to be more effective than pen and paper" (Paulson, 1996).

Since I was pretty sure the advice from an outsider, stranger and above all quite ignorant about cricket would not be received nicely, I kept on sipping tea in my silent corner.

The fascinating evolution of proof tools

"Logic theorist" was the name of the first program developed in 1956 to prove theorems (by the way, the authors were Herbert Simon, the genius inventor of behavioural economics, and one of his colleagues, familiar names in this column).

The procedure followed by the authors of the "logic theorist" consisted in concatenating a collection of coherent algorithms that, given certain rules and axioms, could automatically verify mathematical propositions and come to a conclusion.

It seemed like shaping the magic of artificial intelligence, as every idea can be more or less seen as a concatenation of numbers by mathematical minds. But by the early 1980s the complexity of such theorem proving procedures became a major theme in computer science with a first sense of disillusion and not just because the early theoretical warnings given by the fathers of computer science - Alan Turing and others - were eventually understood.

The assumption of the "logic theorist" was that heuristics for breaking problems into chunks could lead to a chain of assertions and these would be assembled and lead to incontestable mathematical proof. Instead, it turned out there are not so many circumstances involving human reasoning in which things are logic and clear once and for all - and it takes lot of time and energy to normal people to understand it.

Alonzo Church - a guy close to Alan Turing - in 1927 clearly said that there are complex computability problems that cannot be proven by any computer: hence the Church's theorem - one can say... the destiny of a surname!

Anyhow, many researchers and practitioners have been overlooking the Church's theorem for decades and no data scientist complained. It is only a matter of having the right datasets, many keep saying: we have all been accepting this argument as an obvious cover up of insolvable theoretical dilemmas (in Italian, I would say questioni di lana caprina, - issues of goats' hair).

With the birth of enormous repositories of factual and bibliographic data on one side and expert systems and knowledge engineering programmes on the other, the 1960s and 1970s went on investigating the applications of those principles and ideas and anticipated the enthusiasm software entrepreneurs currently show for big data: many funded programmes in artificial intelligence with scope and width not very distant from the ones we have now were devoted to development of proof tools and to the generation of datasets they needed.

Isabelle who?

Progress has been made in many fields of applications of artificial intelligence after that. Proof tools, in spite of some wobbling epistemological foundations, have developed in a successful niche of the STEM markets.

Twenty years after its debut, Isabelle is now one of the most effective proof tools employed by computer scientists, Wikipedia explains, to formalise mathematical theorems that would be extremely difficult if not impossible to demonstrate with pen and paper - for instance, the Gödel's completeness theorem or theorem about the consistency of the axiom of choice, the prime number theorem, the correctness of certain security protocols, and so on and so forth. And there are incredible new developments in the art of verifying and assessing software quality using also other fashionable machine learning methods, less actuarial and more inferential (that means less guessing by associations a la Facebook).

All this is stuff hard to grasp without a degree in physics or mathematics and indeed out of scope here but for reassuring the reader that everybody with average intelligence can approach the meaning of AI concepts in general and understand what the consequences of their applications can be for decision making and human judgement.

For instance, I am learning while writing this article that Isabelle has been used to uncover 144 bugs in an early version of the seL4Kernel of the C Code, and I am sure there are other software and hardware systems in which the tool has been employed with honours.

Considered the level of accuracy achieved in the field of translations and innumerable other common applications, everybody tends to find acceptable that machines are now employed to make or at least deeply orient decisions on matters of law, benefits or any other field where a massive amount of evidence can be processed and analysed automatically.

However, to go back to the point I find intriguing, Larry Paulson warned in 1996 that if it is true that "informal proofs leave big gaps, gaps that minds can bridge but machines often cannot" at the price of leaving us with uncertainty, it is also true that "proof tools require formal calculi, comprising a rigid syntax of formulas and rules for transforming them".

That's not an easy job. In sum, "tools and formal calculi can be hard to use". A degree in computer science or mathematics does not give you, per se, such passport of hard and reliable skills.

Instead, proof tools must embed, again with Paulson words, qualities that we recognise and admire in the human brain independently from what you study or where you come from - although I would say that, of course, studying latin grammar or physics gives you a useful edge:
– Soundness: we can trust the result

– Transparency: we can follow the reasoning
– Power: the tool is mightier than the pen.
There are trade-offs among these benefits. A tool that puts all the emphasis on soundness may sacrifice power, and vice versa. Transparency involves a combination of soundness (the reasoning is correct) and power (the reasoning is expressed at a high level).
Even unsound tools can be valuable: consider floating-point arithmetic. If soundness is not guaranteed then we need transparency, in order to check for ourselves that a derivation is sound.
Soundness can be obtained by recording proofs down to the level of inference rules, and checking them with a separate tool. But this requires considerable storage, and does not aid transparency: detailed proofs are too big to understand. If the proof tool is allowed to invoke external reasoners, such as model checkers or computer algebra systems, then it could record all dependencies on such reasoners.
The tool must let us prove things that we cannot using pen and paper. This is the most important requirement of all, and is perhaps the hardest to attain. Tools are mainly valuable for proofs about objects, such as hardware or formal specifications, that are too big to manage on paper.
(Paulson, 1996)

Mutatis mutandis

I remembered the pub conversation about Isabelle last month, when I received an email from a friend asking me an opinion on the very hot case of the UK Universal Credit programme, the UK major endeavour meant to transform the administration of welfare benefits by means of simplifying the process and realising enormous public spending savings through the application of new digital technologies, including of course artificial intelligence.

My friend was contracted by a legal firm involved in a litigation between the Department for Work and Pension and the family heirs of a claimant who committed suicide earlier this year, while waiting for the Universal Credit claim to be processed. He had a thesis.

He wanted to know if, in my professional opinion and personal experience, any reliable analytics or AI software could help making sense of a large set of unstructured documents to prove the existence of specific policies or intentional conducts on the side of the Government's Department, up to the point that it would cause harm to claimants in order to push them off the welfare system, with devastating consequences for their lives and families. If supported with some strong analytics, such thesis, like a theorem in physics or mathematics, had more chances to be taken seriously in a Court of law.

I am afraid he framed the problem in a way that pointed inevitably towards very subjective, extremely variable, aspects of mental health and fitness but I decided not to argue about his very relevant point of view and I followed his reasoning that was, per se, very intriguing indeed.

I have considered the challenges posed by mobile computing to the legal process and to the law of evidence while I was investigating methodologies to prevent and fight cybercrime in information management few years ago.

Cybercrime is a nested business that pervades several layers of data and information usages not only through straightforward conducts that lead to theft, falsification or fraud by single individuals but also by way of unduly influencing, distorting and providing all sorts of camouflage for the actions of networks of individuals and organised groups in that frustrating the operations of law enforcement teams and the expectations of the same legal and IT communities: "The point rarely taken into account by end users as well as by IT security professionals and computer network scientists - I wrote in that 2013 article, while I was completing a taxonomy of cyber crime - is that there is no certainty about legal directions because of the extremely uncertain relationship between the law and the facts disputed, and not because of a failure of the judiciary system or the law per se in front of the internet or because of an implicit failure of the technology per se neither". (Longo, 2013)

I think I have written about this argument also in several other circumstances: expressing it in clearer or different terms would not make it simpler to understand or circumvent it. So this time around I decided to make a different type of exercise, entering the space of data generation and at the same time make an informative joke for my friend.

I told him that colleagues engineers managed to give me access to a super computer system in Cambridge where they invented a proof tool called Isabelle. So I could have a go at it, asking if a number of statements about the Universal Credit programme - statements that everybody could find reported in academic journals and in the media - were true or false and why, as the proof was the oracle of Delphi! This is actually what people generally imagine an AI tool is for, by the way.

My friend is a lawyer, so he found such explanation not only perfectly sensible but even very functional. To him the difference between a proof tool and an oracle is a quite minimal one: the two are perfectly equivalent, as long as they do the trick or they are admissible as evidence in Court, the Judges do not spot any inconsistency, neither ask to verify their sources, and trust them as expert witnesses' opinions. Who are we all to contradict the insight of modern technological, artificial intelligence magic? The role of computer science in modern life is to give human beings the comfort and healing of religion or hypnosis.

I said to my friend I could not share with him Isabelle's report in its entirety, inclusive of mathematical demonstrations that were copyrighted by the University of Cambridge. But I had made a consistent extract from the lengthy final report. Some responses, interesting and informative, seemed judging the given statements as "true" or "false" at random, in ways not always consistent with the following explanations, in my opinion. But that is the nature of the superior intelligence of all the proof software!, I explained, seriously, to my friend: we have to trust it and take it as it is, including some uncertainties it can manage better than human brains usually do.

Finally, Isabella's conclusions were dragged from an impressive corpus of over 400 sources that I had a chance to see listed: these comprised studies, experts' papers, news, pieces of legislation that the super intelligent software dug up, digested and cross referenced to proof any question statement before giving its output, I lectured. But unlike the few cited in-line, these sources would come in a separate document I had not access to. It would be totally out of scope for the two of us to insist with the engineers to have a copy of the full list of the sources used by Isabelle or its mathematical formulas and demonstrations. But in case he wanted to bring the report in Court, I would do my best to get the whole version including maths and bibliographic references so that he could impress the Judge with that type of well documented, academic research, they all had learned to consider a mark of quality of knowledge during their undergraduate years.

And here we are, what it follows it pretty much what I sent to my friend.

Extracts from Isabelle Final Report on the roll out of Universal Credit for the self employed

Question statement: UC claimants will be always better in work than out of work.
Isabelle response: True. Principle established by the Poor Relief Act (1576): unless disabled, chronically sick, lunatic or blind any claimant able to work must work in a workhouse, otherwise will be put in a house of correction. The modern workhouse is embedded within the Internet: everybody must be "digital by default", set up and maintain a digital twin, find work through the "Find a Job" government portal and made their personal details available for surveillance capitalism.

Question statement: the UC programme was supposed to be signed off at first in 2016-2017.
Isabelle response: True. The implementation of Universal Credit was welcomed with overoptimism and naive trust in a totally unrealistic timescale. The deadline is moving forward every six months and it is now 2023 for a full migration of the 8 million people living in the UK and receiving benefits - the number of which in the meantime will have shrunken by one third. Seen in retrospect, many have found the UC deadline ridiculous because managed through an Agile approach that everybody knows is a method very flexible in many respects but inflexible about deadlines. It seems almost a comedian's trick that the Country that has invented projects controls (with the Prince2 methodology and standard for IT project management) and still sells it all over the World, has embarked for a programme of such scale with an amateurish, "test and learn" Agile approach, pleasing young software developers and IT professionals, academics and quangos, as it was the project of a cheese shop's website. In fact, from that first missed deadline on, criticism from all sorts of stakeholders has flourished and further delays have taken over.

Question statement: the UC programme was a massive welfare reform launched soon after the financial crash in 2009 by the Conservatives' Manifesto, as an electoral manoeuvre meant to radically change the ways in which benefits are claimed, administered and distributed. But its real purpose was to create huge savings in public spending and reduce the amount of benefits that a single household can receive annually, putting a cap to max 23,000£ a year per household.
Isabelle response: True. The idea in itself - that was very successful used by David Cameron, George Osborne and Ian Duncan Smith, together with the less popular ‘bedroom tax’ meant to increase local authorities’ revenues - sprung from the need to contrast and distract the electorate from the harshness of the austerity measures introduced after the 2009 financial crisis. Therefore the programme was strongly intertwined with another meme of the Conservatives' rhetoric at that time that few years later fed into the following Brexit campaign, referendum and debates: the idea of "regaining control of our borders". For several years, the assumption was that with Universal Credit entire segments of the claimants' population would be excluded from any entitlement to benefits and forced into work, bringing to an end a regime of access to welfare considered too soft and generous with several social groups seen as parasites, including EEA nationals, asylum seekers, refugees and the same callous British nationals considered perennials weeds living on benefits, too lazy or too weak to work.

Question statement: the adoption of the Agile methodology for the UC project was motivated by reasons of costs.
Isabelle Response: False. Although in September 2012 Mark Hoban Minister for DWP produced figures for the estimated cost of the UC project of only £638 million, the main attractive reason why Agile methodology seemed fit for purpose at the time was the idea of the fixed, infallible timeframe of development. The new system was due to be implemented in a four years period, between 2013 and 2017. There was an expectation that Agile could succeed in a major IT project where more traditional project methodologies had failed in the previous decade (for instance, the e-Borders project, the central IDs project, the London ambulance computer aided dispatch system project, the Rural payments project, the Integrated Fire Services management project, the NHS for IT project in England and wales, five shared services projects for central government departments and more). Such expectations were widely endorsed by ministers of the coalition Government in 2010-2011 in spite of a warning from the Major Projects Authority in 2011 about the scale of the UC programme.

Question statement: the design of Universal Credit in 2011 was driven by overconfidence and overestimation of the role of shared data among government departments and analytics. Such technology has been heavily marketed by suppliers of IT solutions and outsourcing providers, such as IBM and others, for the last decade. Following the failure and the decommission of a number of major projects in 2010, Ministers endorsed it together with new Agile project management standards. The evidence is that officials and their advisers lack the basic technological and engineering skills to understand what is going on in the labour market.
Isabelle response: False. Real Time Information (RTI) on earnings from employment gathered by HM Revenue and Customs (HMRC) has been used to develop a range of analyses to evaluate earnings and progression of earnings at both individual and aggregate levels. However it is true that the RTI system collects only regular Pay As You Earn (PAYE) information from employers and pension providers. Therefore the shared data coming from real-time flow of information between employers, HMRC and DWP enables Universal Credit to validate earnings every month only for a certain category of claimants.
These shared data also should help the prevention and prosecution of online frauds and tax evasion. But there is some evidence they have been used exactly for the opposite purpose, especially in regard to the management of "advance payments".
Cases of several self employed UC claimants have showed the limits of the exchange of real time information not curated and generated for a great variety of pre-existing and pre-defined purposes other than the administration and auditing of welfare benefits.
Self employed professionals and highly skilled claimants genuinely but not gainfully self employed (on low income) are pushed by the UC Regulations towards low skill jobs (level 1 out of the 4 levels of skills of the UKSOC2010 classification - that comprises jobs such as hotel porters, catering assistants or cleaners) or even to prostitution in order to raise emergency funds and cope with the lack of money for essential expenses for food, clothes and medicines not covered by the NHS. There are also cases of claimants who seek low skills jobs in order to limit their protracted situation of indebtedness.
Such documented cases have shown that both employed and self employed people can have different, and all genuine, income streams, including zero hours contracts, reception of grants and money from charities, declare them all in their self assessment and yet that does not mean they pass a business test or a test of purpose and motivation, neither their earnings qualify as payroll (PAYE) subject to RTI analysis.
The RTI technology is also unlikely to take them into account or be of any significance in a wide range of cases where is impossible to automatically proof whatsoever because of claimants' characteristics and circumstances.

Question statement: DWP needs to ask permission and consent to share claimants' data with external providers and even other government departments.
Isabelle response: False. Data Protection Act 2018 sections 191-194 allows the Government to make totally discretionary use and sharing of people data within Government departments and their suppliers without claimants' consent for a variety of purposes including commercial developments. DWP has recognised in 2018 the complexity around assessing the impact of Universal Credit on the labour market and has decided to work with external organisations to build the evidence base to fully understand what works and how best to maximise labour market outcomes under Universal Credit. This increases the risk that a biased "evidence based" collection of claimants data is under further development. With arguable legal and ethical grounds, claimants have to generate data to populate such collection to comply with their initial UC commitments and profiling. The solution to this circularity problem is unknown at present.

Question statement: the design of Universal Credit has overlooked health issues that affect claimants who may be discriminated in their professional life and have their earnings penalised because of disability, stigma, miscarriages of justice, victimisation, stalking and mobbing, domestic violence or other similar circumstances. For that reason UC does not realise the expected savings on NHS budgets.
Isabelle response: True. The UC modelling relied on assumptions and judgements that underestimated the reality of chronic illnesses, inequality and disability issues while overwhelming relying on the "making work pay" policy. A number of claimants cases demonstrated that positive impacts from possible increases in employment and self employment earnings do not occur because such earnings are in any case capped, conditioned or very precarious, irregular or unhealthy.

Question statement: The Government has embroiled itself in the programme without any evidence they had a viable business case for Universal Credit.
Isabelle response: False. HM Treasury has approved the Universal Credit full business case in 2018, upholding the evidence that it remains deliverable, affordable and provides value for money with a Net Present (social) Value (NPV) of £34 billion over 10 years compared to a total investment cost of only £2 billion. At present, the operational savings per year resulting from changes in the costs of administering Universal Credit compared with the previous system is estimated in £0.3 billion. These figures do not represent the reality. There is no need to represent the reality with UC. Above all, when the Coalition Government started the UC programme in 2011 there was faith in the business case.

Question statement: UC aims at realising savings both in the eligibility and in the administration of welfare benefits and in doing this is punitive and cruel with the "persistently poor" and more vulnerable claimants, exposing them to situations of crisis and other extreme circumstances, in the hope they will turn their lives around, start a business or find employment.
Isabelle response: True. There is no available calculation or study supporting the idea that poorest claimants are receiving less support with UC than with other prior benefits or that there is an embedded cruelty in the design of UC. However, there has been an intense exchange of correspondence between DWP and external providers about the idea of natural migration from other benefits to UC, seen by some as the opportunity to test and innovate the welfare system's strength and by others as a nonsensical and even cruel option pushing vulnerable people towards crisis and suicidal feelings.

Question statement: Both those who designed UC and DWP coaches do not have sophisticated understanding of financial and accounting concepts and procedures and small businesses' practices nor there is any understanding of the role of debt bondages in the labour market, very crucial in cases of domestic violence and modern slavery.
Isabelle Response: False. However, a claimant made the case of inbound cashflow generated through short term loans and charities' donations erroneously considered as earnings by DWP employees, without any further investigation and understanding about the provenience of the money. The same claimant also provided evidence that DWP employees were keen on ignoring money obtained as a grant and as such not to be considered as income or debt generated only because of delays in receiving UC payments. The UC Regulations provide loopholes that are exploited by charities as well as fraudsters and unscrupulous third parties or providers as opportunities for money laundering, modern slavery and human trafficking.

Question statement: under UC, housing benefits are not available anymore for the landlords and / or for others in the local authorities' supply chain because UC is paid only to the claimants that are responsible for their housing costs.
Isabelle Response: False. UC design had to confront a huge corpus of legislation, regulations and traditions that connect health, safety, social risks, housing benefits and local authorities practices. There are still circumstances in which the housing element of UC can be paid directly to the landlord and not to the tenant-claimant. But in general it is true that the monthly UC payment, including the housing element, goes to the claimant. Since the 1890's Housing of the Working Classes Act onwards, local authorities have seen increasing their powers to develop municipal housing and administer social housing benefits until UC has introduced a deep reform of the entire system causing inevitable frictions and conflicts among those who have vested interests in the administration of housing benefits. It has never been an explicit purpose of any welfare benefits legislation including the UC Regulations to support landlords or the "local authorities supply chain". However, it is acknowledged that there is an issue of balance of powers and transformation between central and local government in controlling access to and administration of housing benefits under UC. It is a principle accepted in the whole of the welfare benefits corpus since Malthus's Essay on the Principle of Population (1798) that men should not marry until they could afford to support a family. We as a Country believe in self-help. There are deserving and undeserving poor individuals. Therefore we have always accepted that the homeless individual is less eligible to housing benefits than people with a tenancy agreement in hand, who are also more likely to be in work, or actively looking for work and considered able to work. In 1971 Keith Joseph's government introduced rent allowances and housing benefits (governed by the local authorities) but only to help people in work - and particularly with children and on low wages. Since the Shaftesbury Acts of 1851, local authorities have grown and extended their powers to register, inspect and provide common lodging houses at first and then - also through housing associations - to borrow money at low rates of interest for the provision of working-class housing. This was seen at first as a matter of public health, not as a right for the landlords or for the tenants: public money should be safely used to contrast, prevent and then control the danger of homelessness people and poor people living in slums and spreading infections. The idea was reinforced after Queen Victoria's husband died of typhoid at the age of 42 in 1861, when life expectancy at birth was about 40: the design of a public health system deeply connected with local authorities initiatives and roles emerged in 1866, with the Sanitation Act. The 1868 Artisans' and Labourers' Dwellings Act gave the local authorities also the power to demand property owners to make improvements and to purchase and demolish insanitary properties. (Fraser, 2017)

Question statement: If a UC claimant is considered gainfully self employed he or she is free to develop his or her business or profession, otherwise he or she has to look for a job or start a new business to become entitled to any payment.
Isabelle Response: True. When a UC claimant is considered gainfully self employed he or she is required to develop the business and make it profitable and does not receive any money in excess of what it is deemed to be his or her average monthly salary (also known as Minimum Income Floor) no matter if he or she is able to make a salary at all because this is how the UC mathematical model works. It is also true that in the first roll out of the UC for the self employed, claimants who did not produce enough earnings were asked to close down the business and look for a job or had the option of starting a new business. However this situation did not last for long as a claimant provided evidence that fraudulent businesses could be started fictitiously in order to obtain the new enterprise allowance, just pretending of being non gainfully self employed and in the process of closing down the activity and start a new business.

Question statement: A self employed professional claiming UC is considered not genuine if receives a salary below the minimum threshold of 1100£ per month and therefore treated either as a potential fraudster or as a welfare scrounger. That is because a genuine professional self employed individual is deemed to be obviously profitable and not claiming benefits at all.
Isabelle Response: False. However it is true that stigma and shame have been attached to welfare benefits for long time, particularly in England, causing professionals who receive UC problems of recognition of their competence and expertise, discrimination and unfair treatment in the labour market. If a full time self employed claimant who has a professional status does not obtain earnings above the national minimum threshold, his or her self employment is considered marginal and therefore not genuine and effective, no matter if that is what he or she has been doing for twenty, thirty or forty years. He or she has to look for work, being not gainfully self employed, in order to receive UC. This can create further great disappointment, discontent and even mental health issues among people who have been self employed for a while, who have inherited a family business and other circumstances typical of SMEs and cannot easily adapt to work as catering assistants or "a man and a van" entrepreneur.

Question statement: The formulas used to design UC "making work pay" calculations and mimicking salary payments produce in the end disincentives on taking more working hours and progressing towards better paid jobs for a number of reasons and problems nobody knows how to fix.
Isabelle Response: True. For people in employment, as working hours and earnings rise the combined effects of benefit withdrawal and increased tax and National Insurance liability (measured as the Marginal Deduction Rate, or ‘MDR’) produce negative impacts on workers’ overall income. This, in turn, produces financial disincentives to taking on more working hours and progressing to better paid employment. Worse, it can act as an incentive to leave employment. (Puttick, 2012). For self employed people the negative impact of such disincentives producess loss of productivity and limits their actual potential entrepreneurship. Aware of this type of wicked concerns, in 2016 the Work and Pensions Committee of the House of Commons has launched an inquiry on “in-work progression” for people claiming Universal Credit.

Question statement: Once a claimant's "minimum income floor" (MIF) has been determined that cannot be changed.
Isabelle Response: False. In the first implementation of the UC Regulations for self employed people there was some evidence of friction in the determination of the MIF for self employed people with fluctuation of earnings. The only way to remove it from a claimant's record was to close down the claim, convince the claimant to close down the business or to ask for a mandatory reconsideration that would not be carried forward, with allegations that DWP coaches were instructed to simply ignore complicated cases so that the claimants would go away and look for work. However, in the Summer 2018 a claimant successfully convinced DWP members of staff that such approach was not in the spirit of the UC regulations, a Minister confirmed in a parliamentary debate that more flexibility would be applied in respect of the MIF calculations and DWP members appreciated they have all in all poor maths, legal and business skills and should develop their discretionary role in assessing complex claimants' cases.

Question statement: UC has allowed DWP staff to save time and achieve fairer results with the automation of the Habitual Residence Test (HRT) and the Right to Reside (R2R) Test for EEA nationals claiming benefits.
Isabelle Response: True. Being a de facto permanent or habitual resident in the UK does not mean automatic entitlement to UC for anybody, including British nationals, and therefore such status has to be checked against various Immigrations rules and policies. However, the whole legal matter has been in a state of flux during the inception and then the implementation of UC IT systems between 2010 and 2018 so that the automation of the HRT and the R2R tests have shown misalignment with no fault on the side of the software developers. It is fundamentally inappropriate to say that these developments have been implemented with fairer results for the claimants because they have most of the times played a part as unaware guinea pigs in the IT development and the following changes.
In particular, the EEA Regulations 2006 which transposed Directive 2004/38/EC were initially ignored because the original UC design assumed it would be correct to systematically deny access to benefits to EEA nationals who, after Brexit, would be sent back in their Countries of origin and refused the legal right to reside and treated as all the other foreigners coming into the UK tout court. The right to reside was linked to various requirements such as level of earnings or the status of "qualified person", the combined effect of which would be rejection of EEA nationals as equal citizens with the right to claim benefits even when they had been in the UK for long time or they had acquired the status of de facto permanent residents, with inevitable exclusion and discriminatory treatment in their EU Countries of origin because they had moved their permanent residence and fiscal domicile under the EU Treaty.
The attempt to impose more restrictive British immigration rules to EEA nationals claiming benefits crushed into the reality of the Brexit negotiations and post-Brexit scenarios. The matter went through various legal challenges.
It is now accepted the EU Rights of Residence Directive that has introduced corrections to the freedom of movement dysfunctions. In the UK, under the Immigration (EEA) Regulations 2016, the right to reside was at first granted to EEA nationals according to their employment or self employment status independently from the length of their de facto residence and any other legal status: being a self sufficient person or gainfully self employed was therefore at first the main requirement for any EEA national living in the UK in order to claim UC. On the 21st June 2018 a Claimant made the case to overturn the policy as utterly discriminatory and nonsensical since she had over ten years of continuous residence in the Country. The Claimant successfully faked she was gainfully self employed in order not to compromise her right to reside and human right to keep her habitual residence in the UK, where she came under the freedom of movement / EU treaty in 2008, she was lured for a couple of years into various charitable businesses that caused her to lose over 200K pounds of savings, she started a self employed business in 2011, she was declared bankrupt for non payment of legal costs in a property case she started for a missed conveyance and then restarted a self employed business in 2014. The Claimant's case was brought to the attention of the Home Office that since the 22nd June 2018 has adopted a more balanced and consistent policy towards EEA nationals that have passed the HRT and the R2R tests: these EEA nationals have the right to access the welfare system if they need to, no matter if they are gainfully self employed or not.

Question statement: The UK is the only European Country that has limited access to welfare benefits for migrants from other EU nationals who have less than five years of continued residence.
Isabelle response: False. The limitation of the five years residence has been called "bureaucratic intimidation" and it belongs to a phenomenon observed all throughout Europe for the last two decades, meant to contrast the legal rights of migrants from other EU member states. In the UK, it gets confused also with other issues of conditionality in the tradition of the poor and working class policies and with other recent debated issues about basic income inequality. Access to welfare benefits for migrants and particularly for EU citizens living in another EU member State has been subject to criticism and reforms in several EU countries for years. These measures have had the explicit purpose to restrict, limit or exclude what is known as "welfare tourism"- i.e. foreigners exploiting the freedom of movement to abuse the social insurance systems - and have eventually led to the Directive 2004/38 on the right of citizens of the Union and their family members to move and reside freely within the territory of the Member States [the so called ‘Citizenship Directive’]. Such directive explicitly limits the mobility rights of EU citizens who have less than five years of residence in their Member State of residence (Lafleur, 2018).

Question statement: To claim UC everybody must demonstrate they have acquired the right to reside in the Country with five years of continued residence, no matter their nationality and this process can take up to six months.
Isabelle response: True. However, the UC automated checks system was designed to fight and prevent fraudulent claims so that it discards documents provided to prove the right to reside if these show discontinued addresses of residence, several changes in the number of bank accounts and employers or self employment status, data mismatching with other HMRC data, NHS Business Authority and MoJ records, records held by credit ratings agencies and other preferred sources of information. Claimants with a genuine right to reside who do not pass the automated check need to consider a time frame of about six-months for the assessment of the evidence to be completed manually and go through several R2R tests, self-employment tests and HRT tests.

Question statement: homelessness people cannot make a UC claim because they do not have a postcode or address.
Isabelle response: False. It is true that since 1547, when, after the death of Henry VII, Parliament ordered that slavery was to be the punishment for vagrancy (Fraser, 2017), people that cannot show a tenancy agreement or that they have a landlord are not automatically entitled to the housing component of UC. The Poor Law Act of 1662, aka Act of Settlement, established that a stranger can be removed within 40 days of his arrival in a parish, unless he occupied freehold land, and be sent back to the parish where he came from. There are no relevant updates on leaseholders' entitlement.
Homeless people may still claim assistance and support from their local parishes or local authorities. In fact, UC was designed to support people who have a postcode and address otherwise they should not even start a claim as a measure to prevent and fight welfare tourism. However a claimant made homeless because of a S21 no fault eviction in the process of looking for another rented accommodation and who was not considered by her local parish in priority need - and therefore not entitled to social housing - challenged the UC design stating that until she would find another property to rent her legal address should be considered the one in the Electoral Register, shown on her poll card and that would also be her UC address. The claimant obtained a bereavement payment and other advance payments but the claim was turned down and had to start again several times until the address shown on the claim form matched the one shown on a tenancy agreement and in her record in the Electoral Register.

Question statement: the amount of UC received by any claimant must guarantee a decent standard of living.
Isabelle response: False. UC can never provide total subsistence and the claimants must be always subject to the economy of makeshifts, having to find other means of getting by, find work, start businesses etc etc. In doubt, claimants should always consider having good connections in the local parishes or local authorities, getting along well with relatives and friends, be nice and connect through social media with acquaintances working in financial firms and banks to whom they will provide free publicity on the internet.

Question statement: the inflexibility and rigidity in the design of the UC's IT system emerged as a major problem very early but DWP chose to ignore such warnings.
Isabelle response: True. Departments all throughout the Civil Service are said in demand of much more broad expertise and understanding of changes in policies and the impact these have on operational delivery. A NAO report highlighted in September 2013 how the pilot scheme of UC had revealed lack of capability and flaws in the IT system design, unable to address real changes in claimants' circumstances. The UC design was judged unable to scale and support a national roll out. Also the Major Projects Authority issued a warning in 2013 but DWP insisted for a go ahead, trusting that the adoption of the Agile Government Digital Service "digital by default" approach to interdepartmental projects and data sharing would fix all the design problems improving the IT system day by day.

Question statement: Project management of UC has suffered from cultural misunderstandings, racial divisions and rivalries among ethnic groups involved its design and administration, within the civil service and its suppliers' organisations.
Isabelle response: True. Back in 2010, the Conservatives Coalition Government had promised that large digital and data projects would remain in the UK after various major IT failures, the main cause of which had been identified in the cultural and socio-technical complexity of managing large public contracts with a temporary workforce fragmented, divided by diverse educational and socio-economic backgrounds and afflicted with phenomena of bullying, tribalism and lack of proper skills or experience. Some companies, like HP, had publicly sworn they would never employ people abroad, particularly in India, for major projects development, accepting the evidence of socio-cultural difficulties in implementing social reforms with lack of detailed specifications and offshore contractors. Instead, such promises were forgotten and outsourcing contracts worthing over 500 million each were signed by DWP in December 2011 with Accenture and IBM. These started recruiting IT technicians in Bengaluru and Mumbai. The offshore development and implementation of UC IT system - that at the time was said to be introduced by 2013 - was decided in spite of warnings of risks of errors and large scale, distributed, frauds. Offshoring seemed to some more convenient than opting for massive insurance costs against raised costs of litigations and to others (like the main sponsor of UC, Ian Duncan Smith) a guarantee of time and cost control.

Question statement: The monthly payment of UC, mimicking the real monthly salary, satisfies banks, credit cards, lenders whilst the average UC claimant, especially if poor at budgeting or with very low financial skills, becomes poorer and poorer every day.
Isabelle response: False. This is just an aspect of the policy "making work pay" and "always better off in work". The monthly payment mimicking monthly salary is considered a motivational nudge for the claimants as they are asked for and agree to try to increase their earnings and sources of income in any possible way. On the other side, on a tight low income budget, even with the most spartan lifestyle and without children, claimants are very likely to need additional help from relatives, friends and local communities and rely on short term loans and credit to make a living. Therefore the majority of UC claimants are chronically in debt. They use the money they receive monthly to pay back loans and credit cards or to cover overdrafts and pay interests on various types of private loans. If the claimant has at least basic financial and time management skills and budgeting ability he or she can try to engage in a cashflow management exercise in order to optimise the use of the money and minimise interests, avoiding the risk that the monthly UC payment gets instantly reduced as soon as he or she declares an increase in earnings or income. In any case, the lenders can count on a constant flow of funds to cover small sums of interests regularly paid by UC claimants on short term loans, overdrafts and credit card accounts.

Question statement: The National Audit Office (NAO) employs well paid very highly qualified people to check the consistency and convenience of public spending. In 2013 the NAO highlighted the fact that the UC pilot scheme had shown rigidity and inflexibility of its system and no ability to scale up to the point needed to make a national roll out convenient.
Isabelle response: True. But there is no law or scientific evidence showing that the Government should listen to the National Audit Office or do what they recommend instead of relying on its own internal sources and open data. In any case, the Government could shut down the NAO in order to deliver its policies.

Question statement: The design of UC reveals a poor and stereotyped image and understanding of the job market, aligned with the interests of commercial and financial elites and corporations and large non profit organisations acting cross borders.
Isabelle response: False. Nobody really knows what is going on in the job market and above all nobody within the teams of UC external providers, Ministers or DWP IT delivery offices. The labour market's trends and dynamics remains mostly unrepresented by official statistics. There is an assumption, totally fictitious and / or biased and outdated belief, that employment is the solution to poverty, that there is a job for everybody out there, ignoring phenomena like the commoditisation of skills, the in-work poverty problem, the underemployment, the gig economy and digital slavery trends, the zero hours contracts. Contractors and practitioners involved in the roll out of the UC programme have been overconfidently assuming that the extraordinary fast pace of adoption of ICT by millions or billions of people and the consequent exceptional economic growth of online advertising and retailing for the last two decades can be copied, sic et simpliciter, into other public or private spheres to realise savings or to achieve other ambitious transformative goals, especially in a public sector shaken by austerity measures.
But what works for commercial endeavours of limited scale and no long term plan or for specific policy measures crafted in one culture and political system (eg, India) can be totally unsuitable and unacceptable in other business contexts or geopolitical environments.
The success of Agile and Scrum methodologies for project management, the idealised role of open data and shared data, the enthusiasm of new digital entrepreneurism and innovation policies have coalesced into perfect mechanisms used and abused by Governments' officials to sustain savings and austerity goals and, at the same time, to fuel a narrative of socio technical change and innovation. The two have generated short term private profits and constructed ambiguous policies so that opposite beliefs and biased collections of self-generated data can apparently coexist within the same Government programme in different Departments, all apparently working accordingly to the same new digital and transformative strategies.

Question statement: The government has taken an experimental “test and learn” approach for the implementation of UC and this is the reason why DWP and other departments are not concerned with the critics they received. Instead they see failure as part of the process and tend to self justify themselves for any possible "unintended consequence" of their operations.
Isabelle response: True. However “Test and learn” cannot be a decade-long excuse for failing to properly design a system that is meant to guarantee the social security of so many, and it does not remedy the damage done to those who were thrown into debt or out of their houses, or made to rely on food banks before the improvements kicked in. (Alston, 2018)

Question statement: overwhelming evidence in recent times demonstrates that big reforms require a strategy of long term efforts with plenty of small steps and frequent reviews. Therefore the choice of Agile as a project management methodology was incorrect.
Isabelle response: True. Politicians treat huge programmes as their own short term weaponry to attract and build consensus on major societal change. No matter the methodologies and technicalities, the bigger the change they want to make, the higher are the risks that the projects are actually ambiguous pathways to complexity or chaos anybody is then able to manage according to the ambitions of the initial policies. Major IT projects are therefore a potential source of unintended consequences, uncertainty and dysfunctional costly operations. Contrary to all the expectations and clarity of strategic directions, change simply does not occur in the defined timeframe. The coherent interventions outlined in the business case are not feasible: the problem the programme wants to tackle - it can be the welfare system or climate change, antibiotic resistance or housing, Brexit or cricket's rules - have such a massive number of connections with a cascade of other problems that the impact of uncontrolled changes on businesses, institutions, politics, society as a whole and on the daily lives of millions when not billions of people becomes simply unpredictable and above all unwanted by the majority of people. Change management experts advise therefore to design smaller independent projects along the same strategic directions and implement them with strict controls. This requires not common skills of leadership, ingenuity and management that Ministers very often dislike.

Question statement: Universal credit's experience in the UK demonstrates that poverty, especially since 2010, is a political choice.
Isabelle response: False. Public indignation at the way the UC programme has been implemented is real and has moved the United Nations Rapporteur Philip Alston to talk about Universal Credit with sharp, unambiguous terms within his recent report on extreme poverty and human rights in the UK (Alston, 2018). Few days after the Rapporteur's words, the British Journal of General Practitioners published and article about the "Universal Discredit" echoing, once again, the criticism poured in the public domain from a report the National Audit Office published in June 2018. The NAO focussed on the Universal Credit design and implementation through the "Digital by Default" IT work force, the organisational unit put in charge of the huge digital transformation programme by two Conservative Governments. An article in The Times of the 3rd November 2018 said that the UC programme is a "case of cruelty" and The Sun started a campaign calling for its review. Eventually the same DWP published an internal report, remained buried for three long years, showing that there were reasons for ethical concerns on the first trial of UC sanctions on self employed people on low income that DWP chose to ignore for several years.
In sum, an avalanche of criticism on various aspects of the programme, particularly from 2018 on, has paved the way and eventually pressurised the new DWP Secretary of State Amber Rudd in the Autumn of 2018 to accept the United Nations invite and announce a major review of the implementation plan and the tactical measures, triggering a new parliamentary inquiry on the programme's implementation. In a public statement Amber Rudd has recently admitted that a "deflecting script strategy should not be in place in the future" of UC - that is a partial public admission that a deflecting script strategy has been in place for several years and it is still on the agenda at present.

Question statement: The UC roll out for the self employed has been managed injecting into practices and policies of government's departments a spiral of behavioural changes by financial pressure on the claimants.
Isabelle response: False. This operational style can be seen as a sort of nudging delirium of omnipotence at the root of the UC design, where "digital by default" means actually ground zero in human rights and commoditisation of people skills. Everything is admissible when patronising or even abusing of claimants' time, data, expectations, trust and prospects of success or future options is possible. Claimants have no chances of succeeding nor they can easily legally challenge DWP or provide any evidence of intentional wrongdoing against them. Instead, digital designers have always a chance of demonstrating that they are on their way of reaching the set goals, leveraging on digital technology for good and improving the process, learning lessons from errors etc. etc. The lack of transparency by the UC external providers has been total because outsourced Agile project management work is not documented and above all nobody really controls it. More and more calls for accountability and openly disclosed figures and data have been almost ignored. The NAO has first looked into the UC scheme in 2013, when the press started to report that some IT initial investments would be lost, and yet also another NAO report in 2018 has been mostly ignored. The operations, profits and roles of large contractors involved in the programme, including Accenture, IBM, Capgemini and Hewlett-Packard, have remained mostly undisclosed or obfuscated and unaccounted. The roles of Ministers, Civil Service's Directors, external Consultants, the same Major Projects Authority and even DWP Secretaries of State have been covered by so many people, often with obscure ties to the offshore teams that operate outside the jurisdiction, that nobody seems to have an actual working memory of what has been done and what is going on but the claimants. In a nutshell, it looks like the the main rule of the whole UC programme has been so far "outsource, try again, see if it works and in doubt run away". All in all, that does not mean that "poverty is a political choice" (Alston, 2018).

Conclusions

So what? My friend's reaction to Isabelle notes was, as I was expecting, of great attention and interest but it watered down into a certain blasé attitude when he understood I made up the proof's report and the whole of its background investigation.

If I had not known his way to approach complex and controversial matters of facts, ethics and justice, unlikely to be straightforwardly solved in a Court of Law once and for all, I could have found his conclusion very disappointing. "More or less every claimant gets it in the end" he smiled.

Perhaps in this horribly lighthearted approximation and brushing criticism off the table, there is the proof he was looking for, that human responsibility and judgement in managing programmes of huge social and technological change is always needed. In sum, I am afraid there is not such a thing of positive "digital by default" social transformations. Default is default, mediocrity, failure. Neither the slogan of Government policy professionals "steal with pride" makes any sense at all, but for cybercriminals.

Conversely, in the machismo of the unfortunate Government's slogans, evoking lack of opposition, absence of criticism, exhaustion of other resources for economic growth other than teen-agers' rebellions on social media, in the immature vision and ideology of technological disruptions as expression of a superior creativity of some sort, there is perhaps all the essence and sadness of the big societal failures experienced by digital Britain, from the big society dream to the faith in big data, Brexit and beyond.

Never mind - I said. There are aspects of the cultural approach to the welfare state in Britain so deeply intertwined with being British that I can hardly grasp it, I continued. That is because I am an educated, 1960s generation Italian professional. One of the fundamental Italian constitutional principles and national identity that people of my generation learned in primary school is that working is at the foundation of our collective democratic existence and personal realisation. After more than ten years of full time residence and life in Britain I still struggle to understand the need for a narrative that reminds, top to bottom, "you'll be always better in work than out of work". There is no need to tell me this. Just make it happen.

References

Alston, P. (2018) Statement on Visit to the United Kingdom, by Professor Philip Alston, United Nations Special Rapporteur on extreme poverty and human rights. Retrieved from: https://www.ohchr.org/EN/NewsEvents/Pages/DisplayNews.aspx?NewsID=23881&LangID=E

Lafleur, J-M., Mescoli E. (2018). Creating Undocumented EU Migrants through Welfare: A Conceptualization of Undeserving and Precarious Citizenship. Sociology, 2018, Vol. 52(3) 480–496. Retrieved from: https://www.ncbi.nlm.nih.gov/pubmed/29899582

Longo, B. (2013). Learning on the Wires: BYOD, Embedded Systems, Wireless Technologies and Cybercrime, in "Legal Information Management”, Cambridge University Press, 13, n. 2, p. 119-123. Retrieved from: https://doi.org/10.1017/S1472669613000285

Fraser, D. (2017) The evolution of the British Welfare State. 5.ed. Palgrave.

Paulson, L.C. (1996) Tool Support for Logics of Programs, Computer Laboratory, University of Cambridge. Retrieved from: https://www.cl.cam.ac.uk/techreports/UCAM-CL-TR-406.pdf

Puttick, K. (2012) ‘21st century Welfare’ and universal credit: reconstructing the Wage-Work-Welfare Bargain, Industrial Law Journal, Vol.41, No.2, July, p. 236-249.