The Legal Aid Desert: How a Cut That Booked Its Savings and Never Counted the Cost

There is a number in the records of the family courts that ought to be impossible in a country with a legal aid system, and for nine years it was not only possible but routine. From April 2013 to July 2022, a man accused of beating or raping his partner could stand up in a family courtroom and question her himself, directly, across the well of the court, because neither of them had a lawyer and the law provided for nobody to do it for them. The practice had already been banned in the criminal courts as an obvious cruelty. In the family courts it carried on for the better part of a decade. One woman described to Amnesty International what it meant in practice: facing court meant having to cross-examine the ex she was terrified of. It took until sections 65 and 66 of the Domestic Abuse Act 2021 came into force, on 21 July 2022, for Parliament to stop it - roughly nine years after the cut that created the problem.

That cut has a name, and it is one of the ugliest acronyms in the statute book: LASPO, the Legal Aid, Sentencing and Punishment of Offenders Act 2012. It came into force on 1 April 2013, and what it did was deceptively simple. It did not abolish legal aid. It did something quieter and more durable. It changed the default.

The Inversion
To understand the scale of what happened, you have to understand the single architectural change at the centre of it, because everything else follows from it.

Before LASPO, civil legal aid worked on a presumption of inclusion. Under the Access to Justice Act 1999, a legal problem was within scope and qualified for funding unless it was specifically excluded. The law named what was out; everything else was in. LASPO turned that logic upside down. From April 2013, nothing was funded unless it was listed in Schedule 1 of the Act. The law now named what was in; everything else was out. The House of Commons Library puts the change plainly: where a matter had previously qualified unless excluded, LASPO reversed the position and listed in Schedule 1 those areas that remained in scope.

That sounds like a technicality. It was the opposite of a technicality. By inverting the presumption, the government could remove entire fields of law from public funding not by arguing about each one on its merits but by the simple act of leaving them off a list. Whole categories of what practitioners call "social welfare law" fell out of scope at a stroke: most welfare benefits, most debt, most employment, non-asylum immigration, and most private family law. These are not the exotic edges of the legal system. They are the law as most people actually encounter it - the eviction notice, the benefit sanction, the custody dispute, the spiralling debt, the immigration form filled in wrong.

The stated purpose was to save money. The Ministry of Justice's 2010 consultation, launched against a backdrop of post-crash austerity, projected savings of around £350 million a year by 2014-15. Kenneth Clarke, then Lord Chancellor, made the public case in terms designed to sound like prudence rather than retreat. He pointed out that England and Wales spent an average of £39 per person on legal aid against far smaller sums in France, Spain, Germany and New Zealand, and promised to save the taxpayer £350 million a year by 2015 by ending what he called an overly costly system. He framed it as a retreat from a slide towards a compensation culture, and argued the courts were burdened by the sheer number of people bringing problems before them. The implementation, when it came, fell not to Clarke but to his successor as Lord Chancellor, Chris Grayling - a point worth holding onto, because the question of who owns this policy is one we will come back to.

This is not, fundamentally, a story about whether £39 a head was too much to spend on access to justice. Reasonable people can argue about the right number, and Clarke was not wrong that the figure was high by international standards. It is a story about a government that booked a saving it could measure and never counted - never even tried to count - the costs it had been warned, repeatedly and in writing, would follow. Whether those costs cancelled the saving is something nobody can say, which is itself the indictment. And it is a story about persistence across party lines: the structure was built by a Conservative-led Coalition, but it has now outlived not only its makers but a Labour government elected on a promise of change, and thirteen years on it still stands almost entirely intact.

The Scale of the Cut
Start with the money, because the money is the part the government got right, in the narrow sense that it did what it set out to do.

In real terms, total legal aid spending in England and Wales fell by £728 million between 2012-13 and 2022-23 - from £2,584 million to £1,856 million in constant prices, a reduction of 28 per cent. That figure comes from the National Audit Office's 2024 report, and it is the hardest, most citable number in the whole story. The saving was real. The Treasury got its money.

But the spending figure understates what happened to the actual work, because the collapse in cases was far steeper than the collapse in pounds. The Law Society's analysis of the early-advice tier - "legal help," the first port of call for someone with a legal problem - found that the number of matters dropped from almost a million in 2009/10 to just over 100,000 by 2020/21. That is not a cut. It is a near-disappearance: early legal advice fell to roughly a tenth of its former volume. Put in human terms, that is on the order of 900,000 fewer episodes of early legal help a year than the system delivered before. That figure needs handling with care - a matter start is not the same as a unique person, and some of the fall is scope deliberately removed rather than need going unmet - but it is the closest thing the data offers to a count of the people who once got help at the front door of the system and now do not, and it is worth stating plainly because the supply-side metrics that follow can make a human catastrophe read like an accounting change. The Institute for Fiscal Studies, looking at the immediate aftermath, found that the total number of completed civil legal aid cases dropped by 59 per cent between 2013 and 2016, alongside a real-terms expenditure cut of around 30 per cent over the same window. According to Ministry of Justice statistics, legal help workload now sits at less than a quarter of its pre-LASPO level.

It is worth pausing on why the case numbers fell so much faster than the spending. The answer is that early advice is cheap and the problems it heads off are expensive. A few hundred pounds of legal help at the right moment - a letter to a landlord, a benefits appeal lodged correctly, a debt negotiated - can stop a problem becoming a possession hearing, a homelessness application, a tribunal. Strip out the cheap early intervention and you do not just save its modest cost; you push people downstream into the parts of the system that cost far more, while booking the upstream saving today. We will see this pattern again. It is the engine of the whole story.

What left scope, and what stayed, followed a rough and contestable logic of triage:

Read that list as a citizen rather than a legislator and a pattern emerges. The areas removed are overwhelmingly the legal problems of people who are poor, who are not poor enough to be in crisis yet, and who have no organised lobby. That is not an accident of drafting. It is, as we will see, close to the whole explanation for why the cut generated so little sustained political resistance.

The Geography of the Desert
A right you cannot exercise is not a right. And by the early 2020s, for tens of millions of people, the right to legal aid in whole categories of law had become exactly that - a theoretical entitlement with no provider within reach to deliver it.

The Law Society has mapped this, drawing on the Legal Aid Agency's own directory of contracted providers and Office for National Statistics population data, and the maps gave the phenomenon its name: legal aid deserts. The housing figures track the desert spreading. In 2019, 184 of 348 local authorities had no housing legal aid provider at all, and a further 81 had only one - meaning over half the population of England and Wales lived somewhere with one provider or none. By October 2023, the Society found that 25.9 million people, 43.6 per cent of the population, had no access to a local legal aid provider for housing advice, a rise of 6.6 per cent on 2019. Only a third had access to more than one. Christina Blacklaws, a former president of the Society, put the human geometry of it memorably: in Cornwall, one firm covered a population of over 500,000 across 1,300 square miles, and if your nearest legal aid solicitor is in the next county, they might as well be on Mars.

Housing is not the worst of it. Across other categories, the Law Society's data from 2023 to 2025 found community care deserts covering roughly 70 per cent of the population - over 42 million people with no local provider; education deserts covering 88 per cent; welfare benefits, 82 per cent; immigration and asylum, 63 per cent.

Behind the deserts is a simple collapse in the number of firms and agencies willing to do the work. Providers holding civil legal aid contracts fell from 2,129 in April 2012 to 1,401 by 2023; the IFS frames the longer arc as a fall from roughly 3,000 in 2010 to around 1,400 by 2024 - more than a halving in a decade and a half. The Law Society states that the number of advice agencies and law centres doing legal aid work fell by 59 per cent compared with 2012. The NAO found that the proportion of the population within ten kilometres of a housing legal aid office fell from 73 per cent to 64 per cent across the decade.

The cause is not mysterious, and it is the second structural fault at the heart of this story. Civil legal aid fees have not risen since 1996, and were cut by 10 per cent in 2011-12, which leaves them at roughly half their 1996 value in real terms, according to the NAO's 2024 analysis. Thirty years of inflation have done to the fee what a single dramatic cut never could: they have made the work uneconomic to do. A firm cannot run a legal aid department at a structural loss out of charitable feeling indefinitely, and most have stopped trying. Research by Dr Jo Wilding of the University of Sussex, using freedom-of-information data, found the consequence in stark terms: in nine of the 131 housing procurement areas, not a single new housing case was opened in a year; in welfare benefits, 71 per cent of offices reported no new legal aid cases at all; and in asylum there was at least a 45 per cent gap between demand and provider capacity, with at least 25,000 more people seeking legal aid than providers could take on.

That last set of figures is the desert made literal. A scheme can remain perfectly intact on paper - the entitlement listed in Schedule 1, the contract available to tender for - while delivering nothing, because nobody can afford to pick up the work at the price on offer. The entitlement survives. The advice does not.

The Litigants Left Holding Their Own Cases
When you take the lawyers out of a system built around lawyers, the cases do not vanish. The people still come to court. They simply come without representation, and the courts have spent a decade absorbing the consequences.

In the year after LASPO, the NAO recorded a 30 per cent year-on-year increase in family court cases in which neither party had legal representation, and a 22 per cent rise in cases involving contact with children. A decade on, the new normal had set hard: by the first quarter of 2023, in 40 per cent of family dispute cases neither party had representation, a 26 per cent increase on ten years earlier. The Nuffield Family Justice Observatory reported in 2024 that around 80 per cent of private family law cases now involve at least one party without a lawyer.

Sir James Munby, the former President of the Family Division and no one's idea of a firebrand, has been about as scathing as a senior judge permits himself. He described the effects of LASPO as baleful and, he feared, all too predictable - indeed actually predicted - and said the Act had made the family court an increasingly lawyer-free zone, with ever more litigants appearing unrepresented and without advice. He called the government's accompanying push towards mediation a disastrous failure, and warned in 2020 that the private family law system, under heavy and sustained attack, risked being brought to its knees.

The official research bears out that this is not a matter of inconvenience but of justice miscarried. The Ministry of Justice's own work found that the major reason for appearing in person was inability to afford a lawyer; that only a small minority of litigants in person could represent themselves competently; and that around half of those observed had one or more vulnerabilities. These are, very often, people in the most consequential disputes of their lives - whether they will keep their children, their home - navigating procedure and evidence and cross-examination with no idea how any of it works, against, sometimes, a represented opponent who does.

And here the cost-shift shows itself with particular clarity, because litigants in person do not save the system money. They cost it. Cases involving unrepresented parties can take up to 50 per cent longer, as judges stop to explain procedure, untangle irrelevant material, and do work a lawyer would have done outside the courtroom. The NAO estimated additional cost to the courts service of at least £3 million a year, with the family-court litigant-in-person increase alone estimated to have cost £3.4 million in 2013-14. The Ministry committed around £2 million over two years for litigant-in-person support - less than the annual cost it had created. The saving on legal aid was booked at the Ministry of Justice; part of the cost reappeared, unbooked, a few doors down the same building, in the courts budget. This is cost-shifting in its simplest form, and one of the few corners of the system where it can be seen with a figure on both sides: not the disappearance of a cost, but its relocation to a column nobody was adding up. Hold on to that, because the aggregate version of the same manoeuvre - across the NHS, local authorities and the rest - is precisely the thing the government never measured.

The cross-examination scandal with which this article opened belongs here too, as the sharpest illustration of what "unrepresented" can mean. Because LASPO had stripped legal aid from private family cases, alleged abuser and alleged victim alike could find themselves in court without lawyers, and the alleged abuser could question the victim directly - a practice banned in the criminal courts precisely because of its cruelty. Sections 65 and 66 of the Domestic Abuse Act 2021 finally prohibited it, inserting provisions into earlier legislation and providing for court-appointed Qualified Legal Representatives paid from central funds. They took effect on 21 July 2022. But the fix has its own shortage: there are too few practitioners willing to take the QLR work at the rate offered, a problem the President of the Family Division acknowledged in a 2024 judgment dealing with a case where no QLR could be found. The withdrawal of funding produced this cruelty - foreseeably, and on Munby's account actually foreseen - the state then took nine years to ban it, and then under-resourced the remedy so that it sometimes cannot be delivered at all. It is the same shape of failure that recurs throughout this story: a harm the policy caused, a fix arriving late, and the fix itself starved of the means to work.

The Safety Net That Caught Almost Nobody
Every account of LASPO from its defenders comes back, sooner or later, to the same reassurance: there was a safety net. Exceptional Case Funding, written into section 10 of the Act, was meant to catch the cases that fell out of scope but where refusing legal aid would breach a person's human rights or EU law rights. Lord McNally, the minister who took the Bill through the Lords, sold it as exactly that - a safety net to ensure the most vulnerable would not be denied access to justice. The promise of the net was, politically, what made the removal of scope tolerable.

So it matters a great deal how the net actually performed, and the answer is that in its first year it caught almost nobody.

The government's own pre-LASPO modelling, as later recorded by Mr Justice Collins, had estimated 5,000 to 7,000 ECF applications a year with around 3,700 grants. The reality, in the scheme's first year from April 2013 to March 2014: just 1,519 applications were made in the first place - a quarter of the lower forecast - and of the 1,083 then determined, only 35 were granted. A grant rate of around 3 per cent. And of those 35 grants, 21 were for inquests, leaving just 14 for every other area of law combined - 8 in family, 3 in immigration. Strip out the inquests and the success rate was approximately 1 per cent. The net designed to catch thousands caught, in the categories it most needed to cover, a few dozen.

It was not that the cases did not exist. It was that the scheme had been built to be nearly impossible to use - the forms forbidding, the test opaque, the process designed in a way that defeated the very people it was meant to protect, most of whom were applying without the lawyer they were applying to fund. In July 2015, in IS v Director of Legal Aid Casework, Mr Justice Collins declared the scheme's operation unlawful and, in his word, wholly deficient. Litigation gradually forced it open. Grant rates climbed - to around 65 per cent by late 2018 after a series of cases including Gudanaviciene. But the recovery is its own indictment. A Public Law Project survey found that 77 per cent of providers still did not regard ECF as effective, with 61 per cent strongly disagreeing. And the scheme now consumes an absurdity: the Ministry of Justice spends nearly £400,000 a year assessing immigration ECF applications, of which almost 90 per cent are approved. For that category, at least, the gate has stopped doing the job a gate is for. When nearly nine in ten applicants are waved through at a cost of six figures a year, the assessment is not protecting the public purse; it is an expensive turnstile in front of a door that is effectively open. Across the scheme as a whole the picture is less extreme - grant rates run at around 65 per cent, so a third of applicants are still refused, and the gate is not a sham everywhere. But it is precisely in immigration, the largest category, that the cost of assessing has come adrift from any gatekeeping it actually performs. The rational thing would be to bring those cases back into scope and stop paying to assess them. That the government has not done so tells you how much the architecture of LASPO is now defended for its own sake.

The mandatory telephone gateway was the same story in a different key. LASPO required that debt, discrimination and special educational needs matters be accessed first through a Civil Legal Advice phone line. The Ministry had believed over half a million cases a year could be handled by phone, with 5 to 10 per cent referred onward for face-to-face advice. It was drastically underused. In 2016/17, exactly one client received face-to-face education advice through the gateway; not one received face-to-face discrimination advice. The Justice Committee found that the failure to provide adequate public information was a primary reason the gateway went unused, and that the underuse was a primary reason for an 85 per cent shortfall in debt cases. The mandatory gateway was eventually phased out. Another safety mechanism that existed mainly on paper.

A Reckless Economy, Documented Twice
The strongest part of the case against LASPO is not an opinion. It is the considered verdict of the National Audit Office, delivered once in 2014 and again, in almost identical terms, ten years later in 2024 - bracketed by two reports from the Public Accounts Committee saying the same thing. The state's own auditors examined the policy and concluded, twice, a decade apart, that the government had never understood the costs it was shifting onto the rest of the public sector.

The 2014 NAO report, Implementing reforms to civil legal aid, found the Ministry on track for its savings but unable to demonstrate value for money. Its head, Amyas Morse, put it precisely: without an understanding of the wider costs, the Ministry's reforms could not be said to have delivered better overall value for money for the taxpayer. The report documented the mechanism failing in real time. The Ministry had expected family mediation referrals to rise by 9,000 a year once people were steered away from court; instead there were 17,246 fewer mediation assessments in 2013-14, a 56 per cent fall. People did not mediate more. They simply went to court alone, or did nothing. In 14 local authorities, no provider began any face-to-face legal aid work at all in 2013-14, and the Legal Aid Agency, in the NAO's words, did not know why and so could not be confident it was targeting funding at those most in need.

The Public Accounts Committee, then chaired by Margaret Hodge, was blunter still in its February 2015 report. The reforms, it concluded, were based not on evidence but on an objective to cut costs as quickly as possible, and the Ministry did not know, and had shown little interest in, the knock-on costs across the wider public sector. The Permanent Secretary had told the Committee that the level of spend was the critical factor driving the reforms - which is to say the Ministry had been candid that the point was the saving, not the system.

The remarkable thing - the thing that turns this from a policy misjudgment into something closer to a structural refusal to learn - is that ten years later, nothing had changed in the Ministry's grasp of its own policy. The 2024 NAO report found that the Ministry still lacked an understanding of the full costs and benefits of its reforms. The accompanying PAC report, in May 2024, found the Ministry still could not demonstrate overall savings to the public purse, and that its own survey showed providers turning away an average of 26 eligible cases each per month. A decade of being told to collect the downstream-cost data, and the data still did not exist.

What were those downstream costs? The honest answer is that the government never properly measured them, which is itself the charge. But the estimates that exist all point one way. Pressure rebounded onto the courts, the NHS, local authorities, prisons, school exclusion processes and social services. A survey found that half of respondents eligible for legal aid said their unresolved civil legal problem had worsened their health and wellbeing - and ill health has a cost, borne by the NHS rather than the Ministry of Justice. Citizens Advice calculated a return of £2.34 for every £1 spent on housing legal advice. The loss of VAT alone on legal aid work was estimated at £60 million for 2013-14. And crucially, the warning had been on the record before the Act passed: a 2011 report by Dr Graham Cookson, Unintended Consequences, had estimated that the projected £270 million saving would be offset by at least £139 million in knock-on costs to other public bodies. The Ministry did not test that warning. It legislated past it.

Here a note of fairness is owed, and the discipline of the argument requires it. Several of the downstream-cost estimates - Citizens Advice, the Cookson report, various charities - come from organisations with a stake in the conclusion. They are advocacy figures, and should be read as estimates rather than audited fact. But that caveat cuts both ways, and mostly against the government. The reason the only available figures come from advocates is that the government, having been told by its own auditors twice to produce robust data on wider costs, never did. When the only people who have bothered to count the costs are the people campaigning against the cut, that is not a reason to dismiss the count. It is a measure of the Ministry's incuriosity about the consequences of its own policy.

And here honesty requires me to draw a line I have so far let blur, because it is the line on which the whole "false economy" charge turns. There are two different versions of that charge, and only one of them is proven. The weaker version is that the government booked a saving it could measure while refusing to measure the costs it created, so that nobody - the Ministry least of all - actually knows whether LASPO saved public money once everything is added up. That version is established beyond serious dispute, by the government's own auditors, twice, a decade apart. The stronger version - that LASPO demonstrably cost more than it saved - is not established, and a careful reader should notice that my own best figure cuts against it. Cookson's estimate was a £270 million saving offset by at least £139 million of knock-on costs: an erosion of roughly half, not an erasure, and on its face still a net saving. The "at least" is doing real work - the costs Cookson could not quantify might well close the gap or cross it - but I cannot assert that they do, because the data that would settle it does not exist. That is, again, the whole point. So let me state the thesis at the strength the evidence actually supports: not that LASPO lost money, but that it may well have, that its design made that outcome entirely plausible, and that a government seriously interested in value for money would have found out rather than spending ten years declining to. A false economy in the strict sense - a cut that cost more than it saved - remains unproven. A reckless economy - a cut taken without ever establishing whether it paid - is proven to the hilt, and is the more damning charge precisely because the government had every opportunity to rebut it and never tried.

What Reform Looked Like, and Why It Didn't Happen
It is not as though no one drew up a plan. The most serious attempt came from the Bach Commission on Access to Justice, established in 2015 with Fabian Society support and chaired by Lord Willy Bach - a former legal aid barrister and justice minister - with the retired Lord Justice of Appeal Sir Henry Brooke as his vice-chair. It took written evidence from over 100 individuals and organisations and reported in September 2017 under the title The Right to Justice.

Its centrepiece was a proposed Right to Justice Act, which would have codified a legally enforceable right to reasonable legal assistance, overseen by an independent body, restored early legal help to pre-LASPO levels across social welfare law, scrapped the mandatory telephone gateway, and replaced the Legal Aid Agency with an arm's-length commission. The estimated cost of restoration, cited in the December 2017 Lords debate, was around £400 million - which is to say, roughly the saving LASPO had made, give or take, before any account of the downstream costs that saving had generated. The shadow justice secretary at the time, Richard Burgon, said the recommendations would form an important part of the next Labour manifesto. The justice minister Dominic Raab was dismissive, noting that the government had spent £1.6 billion on legal aid the previous year, a quarter of the Ministry's budget, and characterising Labour's proposals as unfunded.

The government's own stocktake, the Post-Implementation Review of Part 1 of LASPO, came in February 2019. Its conclusion was a study in the passive mood: LASPO had been partially successful, in that savings had been achieved, but it was impossible to say whether it had discouraged unnecessary litigation or targeted those most in need - the two things, other than saving money, it had been sold as doing. The accompanying Legal Support Action Plan committed to pilots and modest sums: up to £5 million for innovation, £3 million over two years for litigants in person, the reintroduction of face-to-face advice in a few categories, and a pilot of early advice. The Bar Council's then-chair called it a drop in the ocean. The Justice Committee chair, Bob Neill, warned that the proliferation of proposals for further review risked being seen as kicking the can down the road. Crucially, as the review itself acknowledged, it did not fundamentally change LASPO as the basis for providing legal aid. The architecture stood.

Some genuine, if narrow, restorations followed over the next few years: special guardianship orders brought back into scope from May 2023, legal aid for separated migrant children, and a revision of the means-test thresholds in 2023 that was estimated to make over two million more people eligible, at a cost of around £25 million a year. These are real improvements and should be recorded as such. But they are repairs to the fittings, not to the foundation. The inverted presumption of scope - the thing that defines LASPO - remained untouched.

The Government That Was Going to Be Different
In July 2024, Labour won a landslide. Here, if anywhere, the structure built by the Coalition might have been expected to come down, or at least to be substantially rebuilt. The party had commissioned the Bach Commission. Its conference had heard the case for restoration. Its lawyers and its MPs - particularly those with heavy welfare caseloads - had spent a decade watching the consequences arrive in their constituency surgeries.

What has actually happened is modest to the point of being hard to characterise as restoration at all. Shabana Mahmood served as Lord Chancellor from July 2024 - the first Muslim and the second woman formally appointed to the office (after Liz Truss; some counts reach back to Eleanor of Provence, who wielded the powers in 1253 without ever being appointed, and call Mahmood the third) - until September 2025, when she moved to the Home Office and David Lammy replaced her, becoming the tenth Lord Chancellor in a decade. On civil legal aid, the government has restored no scope. Its concrete action has been fee uplifts, and only in two areas. In November 2024 Mahmood announced a minimum 10 per cent uplift for housing and immigration legal aid fees, the package worth around £20 million a year by 2027-28, and described by the government as the first meaningful increase in civil legal aid fees in almost 30 years - a phrase that is also a quiet confession of how long the fees had been frozen. The final regulations, laid after a judicial review brought by Duncan Lewis Solicitors, went further for the most loss-making work: the Civil Legal Aid (Procedure and Remuneration) (Amendment) Regulations 2025 delivered effective Controlled Work uplifts averaging around 42 per cent for housing and 31 per cent for immigration and asylum, the exact figure varying with a provider's case mix, and took the London immigration hourly rate to £69.30 - which, the firm noted when the rate was confirmed in July 2025, still leaves it nearly 50 per cent below its 1996 value in real terms. The uplifts came into force on 22 December 2025.

These are not nothing. An uplift of that size on rates frozen for three decades will keep alive some housing and immigration practices that would otherwise have closed, and that matters directly for the deserts. But notice the shape of what has been done. The other civil categories - mental health, education, community care, discrimination - received no uplift and remain, in the government's phrase, under consideration. The increase does not apply to cases opened before 22 December 2025, a decision Duncan Lewis criticised as likely to slow any recovery in capacity, since providers keep being paid the old unsustainable rates for all ongoing work. And a fee uplift, however welcome, is a repair to the second structural fault - the frozen fee that emptied the deserts - while leaving the first, the inverted presumption of scope, entirely in place. A better-paid lawyer is no use to a citizen whose legal problem is no longer in scope at all.

The government's Review of Civil Legal Aid, launched and reporting its evidence base across 2024 and 2025, with a consultation that closed in March 2025, may yet propose something more structural. But at the time of writing it has not, and the Law Society's verdict on the state of the system - delivered by its then vice-president in response to the 2024 NAO report, that both the civil and criminal systems are on their knees - still stands. Thirteen years after LASPO came into force, under a government that ran against the politics that produced it, the inverted presumption survives. The desert has now outlasted the Coalition that made it, two Conservative successors, and the first eighteen months of a Labour government elected to be different.

This is the point at which the question of blame has to be faced squarely, because it is tempting to make this a story about one party. It is not. The Coalition designed and implemented LASPO - Clarke architected it, Grayling delivered it. But Labour, having commissioned a detailed and costed plan for restoration in opposition, has in power restored no scope and delivered only targeted fee uplifts. Both main parties have now had the chance to dismantle the structure, with full knowledge of its effects, documented by their own auditors. Both have declined. The bipartisan character of the failure is not a rhetorical flourish. It is the most important single fact about why the desert persists.

The International Picture, Carefully Stated
There is a claim that gets made about all this, and it is worth handling with precision, because the loose version of it is wrong and the precise version is damning.

The loose version is that England and Wales now has one of the worst legal aid systems in the developed world. That is not true, and saying it discredits the rest of the case. The precise version is that England and Wales underwent one of the largest retrenchments of civil legal aid scope of any comparable wealthy democracy in the 2010s - which is true, and is something quite different.

The distinction turns on the baseline. England and Wales started from an exceptionally high position. The Council of Europe's CEPEJ data for 2012 placed it third in Europe at €41.55 per inhabitant, behind only Norway and Northern Ireland, against a Council of Europe median of about €2.3 - a figure so low that fifteen countries spent under €1 a head. So when LASPO cut deeply, it cut from a great height, and even afterwards England and Wales remained near the top of the European table in absolute per-capita terms. The CEPEJ figure for England and Wales fell from €38.14 in 2014 to €22.25 in 2020 - a decline of around 42 per cent over six years, and the cleanest single illustration of the post-LASPO trajectory - but CEPEJ's own commentary noted that even at €22.25 the figure remained around seven times the European median. You cannot honestly say a country spending seven times the median has the least access in Europe.

So the defensible claim is about the size of the cut, not the lowness of the floor. And on that, the comparison with the rest of the United Kingdom is the most telling, because it holds more constant than any cross-Channel comparison can. Scotland, sharing a state and a Treasury cycle, did not replicate LASPO. Under the Legal Aid (Scotland) Act 1986 it retained the broad-scope model - everything in unless specifically excluded, the exclusions being narrow - and focused its reforms mainly on criminal legal aid. The House of Commons Library confirms that the biggest changes occurred in England and Wales. Scotland also kept markedly more generous eligibility thresholds.

I should resist the temptation to call this a natural experiment in which only the policy differs, because it plainly is not: Scotland has its own legal system, its own profession and court structures, a devolved government answering to a different electorate under different incentives, and a smaller and differently distributed population - any of which shapes how legal aid is delivered and how much it costs. The same scruple that makes me reject "worst in the developed world" two paragraphs ago forbids me the tidy "controlled experiment" here. What Scotland does control for is the single variable the "this was just austerity, everyone was cutting" defence rests on: it faced the same post-2010 fiscal squeeze from the same Treasury, and it did not invert its presumption of scope. That makes the divergence strong evidence that England and Wales's retrenchment was a policy choice rather than a fiscal inevitability. It is not laboratory proof, and I will not dress it up as such. It is the best available comparison, which is a different and more modest thing.

The Netherlands is the other instructive comparator, because it shows a different path not taken. Dutch attempts at aggressive scope-cutting were defeated in the Senate in 2015; the later reforms from 2018 focused on how providers were paid and how services were structured, against a backdrop of fee freezes and provider attrition, rather than wholesale removal of scope. The Dutch squeezed the supply side, much as England's frozen fees did, but did not perform the legislative inversion that defines LASPO.

One genuine caveat governs all of these comparisons, and a careful account has to state it rather than bury it. Legal aid figures are not like-for-like across legal systems. In adversarial common-law systems such as England's, the cost of building a case falls on lawyers, which inflates the legal aid budget; in the inquisitorial civil-law systems common on the Continent, much of that work falls on judges and courts, and shows up in a different budget line. CEPEJ itself warns that comparing is not ranking, and that its figures must be read with the accompanying metadata. The Ministry of Justice's own commissioned research once noted that while legal aid costs in England and Wales were unusually high, spending on courts and prosecution was comparatively low - the same work, paid for in a different column. This is why the per-capita league tables should be treated as directional rather than precise, and why the strongest version of the international argument rests not on the absolute figures but on the within-UK comparison, where the same fiscal pressure produced a different policy choice.

Why So Quiet?
A cut of this scale, with effects this documented, might have been expected to generate sustained political resistance. It did not. The protest came largely from the legal profession and the advice sector - interested parties, easily discounted as such - and from a minority of MPs. The wider public mostly did not notice, and the reason is the most uncomfortable part of the whole story.

The areas LASPO removed from scope were the legal problems of the poor, the precarious and the marginal: people facing benefit sanctions, debt, eviction at the early stage, immigration difficulties, family breakdown without the funds to litigate. Professor Ben Bowling of King's College London put the political economy of it plainly: allowing ordinary people to seek redress in court is a way of defending the poor, and that is not a vote winner; some of the groups worst affected - young people in care, the homeless, migrants - are also among society's most voiceless. These are people who are less likely to vote, less likely to be organised, and who lack any natural constituency to carry their grievance into Parliament. A cut that fell on them could be made with little electoral risk, which is close to a complete explanation of why it was made, and why it has been so hard to reverse.

The one channel through which the cost did rebound on politicians is revealing precisely because it is the exception. When free legal advice disappeared, some of the displaced demand landed on MPs' own constituency surgeries, and research found that the cuts shifted casework onto MPs' offices. The most concerted parliamentary disquiet came from Labour MPs with heavy welfare caseloads - the ones for whom the downstream cost arrived, personally, in the form of desperate constituents with nowhere else to turn. The system has a way of making its costs felt eventually. It just routes them away from the people who designed it and towards the people, and the institutions, least able to refuse them.

The human texture of those costs is documented, for anyone willing to look. Amnesty International's report described people with mental health problems at risk of homelessness, parents fighting for access to their children, and migrants left unrepresented, and warned of a two-tier civil justice system - open to those who can afford it, increasingly closed to the poorest and most in need of its protection. The Equality and Human Rights Commission's study of the Liverpool City Region interviewed over 100 people and found that LASPO had forced many to take no action at all, because the advice they needed was no longer free and no longer there; it found, among other things, that initial legal aid for discrimination cases dropped by nearly 60 per cent after the telephone gateway was introduced. The New Statesman documented a sixteen-year-old victim of sexual abuse in local authority care among those left without representation. These are not edge cases dredged up for effect. They are what the routine operation of the policy produces, and they are the reason the abstraction "scope" is worth puncturing wherever possible.

The Case the Reformers Could Have Made
The argument I have built treats LASPO as a cut wearing the costume of a reform. Fairness requires admitting that there was a real reform underneath the costume, and that some serious people - including some who came to detest what LASPO became - believed in it. A piece that only ever defeats the weakest version of the opposing case has not really won the argument; it has avoided it.

So here is the strongest version, and it has nothing to do with saving money. A legal aid system that funds a lawyer for almost any civil problem, and pays that lawyer to litigate, is not self-evidently the best way to help people with civil problems. It is lawyer-centric by design, and a system that pays lawyers to litigate will, at the margin, produce litigation. A good many of the disputes that filled the pre-LASPO caseload - separating couples fighting over contact, neighbours at war, a benefit refusal that turns on a misread form - are not obviously best resolved by two sets of publicly funded solicitors and a contested hearing. The intuition that adversarial litigation is a poor first resort for a separating family is not a Treasury talking point. It is held by family-justice reformers who loathe LASPO, by senior judges, and by the mediation profession whose entire premise is that courts should be the last resort and not the first. The case for steering such disputes towards mediation, early advice and non-legal support, and away from the court door, is a respectable one, and it predates by years the austerity that later hijacked it.

On that reading, LASPO was meant to do three defensible things at once: stop publicly funding the litigation of disputes better handled another way; redirect people towards cheaper and less adversarial forms of help; and concentrate the money that remained on the cases where liberty, safety or the roof over someone's head was genuinely at stake - which is why asylum, detention, domestic-violence family cases and homelessness stayed in scope while general advice and private family disputes fell out. Read charitably, the scope list is not random cruelty. It is an attempt, however crude, to draw a line around the cases that most need a lawyer and to deal with the rest some other way. An informed defender of legal aid reform would recognise their position in that paragraph, which is the test of whether the steelman is honest.

None of this contradicts the earlier point that the cut fell on the voiceless. Both things are true at once: the triage had a defensible logic, and the fact that the line it drew fell on people without an organised lobby is exactly why a contestable rationale met so little resistance. A policy can have a real argument behind it and still survive mainly because its losers could not fight back.

I think the case fails - but it fails for a specific reason worth stating precisely, because the reason is not the lazy one that reform is always a cover for cuts. It fails because the reform half was never actually built. A genuine version of this policy would have funded the alternatives it was steering people towards before it withdrew the lawyers: resourced mediation properly, expanded early advice and non-legal support, made the telephone gateway usable. Instead the government removed the funded option and assumed the alternatives would materialise of their own accord. They did not. Mediation assessments fell by 56 per cent rather than rising; the gateway went all but unused; and the early-advice tier that was supposed to catch problems upstream was itself gutted, falling to a tenth of its former volume. You cannot claim to be redirecting people to better channels while simultaneously defunding the channels.

I should mark the limit of that rebuttal, though, because it does not bite equally on everything LASPO removed. Where there genuinely was a better channel - the separating couple better served by mediation than by a contested hearing, the benefits problem better solved by sound early advice than by a tribunal - defunding the lawyer while failing to fund the alternative is precisely the contradiction just described. But for some of what was removed, a defender can fairly say no publicly funded resolution was warranted in the first place, by lawyer or by anything else: the low-value private dispute the state has no strong reason to bankroll at all. For those categories the removal may stand on its own merits, alternatives or no, and I will not pretend otherwise. The "you cannot defund the channels" argument is decisive only where a better channel actually existed and was then starved - which covers most of social welfare law, but not quite all of what LASPO touched. That is the honest boundary of my own case, and it is worth marking precisely so that the part inside it can be pressed without apology.

The Permanent Secretary's admission that "the level of spend" was the critical factor is, on the fairer reading, not proof that there was no reform theory at all, but proof of which half of the theory the Ministry actually cared about. The reform was real on paper. On the ground, only the cut was implemented - and across the wide range of cases where a better channel did exist and was defunded anyway, a reform delivered as a cut is, in its effects, just a cut with a better explanation.

The Choice That Has Been Deferred
Step back, and the structure of the thing is clear.

LASPO did two things at once, and both have to be understood to see why the desert persists. It inverted the presumption of scope, removing whole fields of law from funding by leaving them off a list - and that inversion remains in force today, untouched by thirteen years of reviews, by the Bach Commission's costed plan, by two changes of governing party. And it sat that inversion on top of a fee that has been frozen since 1996 and is now worth half what it was, which has hollowed out the providers and turned the surviving entitlements into deserts where no one can afford to do the work. The Labour government's fee uplifts of December 2025 begin, tentatively and in only two categories, to address the second fault. Nothing has yet addressed the first.

And running underneath both is the verdict the National Audit Office reached twice, a decade apart: that the Ministry pursued its savings without ever establishing what they cost the courts, the NHS, local authorities, and the people themselves - and did not go looking, even when its own auditors twice told it to. The Ministry was warned before it acted, by Dr Cookson, that the saving would be substantially offset. It was told in 2014, by the NAO, that it did not understand the wider costs. It was told again, in 2024, that it still did not. The defining feature of the policy is not that it was cruel, though in the family courts it was, for nine years, exactly that. Nor is it that it lost money, which cannot be shown either way. It is that the government undertook a structural change to access to justice without ever establishing whether it left the public better or worse off in the round, and that ten years of being asked to find out produced no answer - because the question was never one the Ministry wanted answered.

There is a version of the response to all this that says: spend more. It is not wrong, but it is incomplete, because money poured into the existing structure would partly drain away through the same holes - into a scheme like ECF that costs £400,000 to wave through 90 per cent of applicants, or into fees so low that providers still will not do the work. The prior question is structural. Do you restore the presumption of inclusion, bring early legal help back into scope for social welfare law, and price the work so that someone will actually do it - accepting the cost, and counting honestly against it the downstream savings the government has never bothered to measure? Or do you keep the inverted default, keep the deserts, and let the two-tier system that Amnesty described harden into permanence: justice as a service for those who can pay for it, and for those who cannot, a litigant-in-person form, a phone line nobody answers, and a courtroom they must walk into alone?

That is a political choice, and it has been deferred - through a post-implementation review, a costed commission, a string of pilots, two damning audits, and now a third review still running - for thirteen years. The benchmarks that would show it had finally been made are not mysterious. Early legal help restored for social welfare law to something near its pre-LASPO level. Fee uplifts extended beyond housing and immigration to every civil category. A statutory access-to-justice duty of the kind the Public Accounts Committee called for. And the Ministry of Justice, at long last, collecting the downstream-cost data that its own auditors have demanded for a decade. Until those things happen, the structure of LASPO stands, the desert remains on the map, and the verdict the National Audit Office reached in 2014 and confirmed in 2024 remains exactly where it was: a saving counted to the last million, costs the government was warned of and chose never to weigh, and no one able to say whether the country came out ahead - because no one with the power to find out has been willing to.

This piece draws on a verified fact base assembled from the National Audit Office (2014 and 2024 reports), the Public Accounts Committee (2015 and 2024 reports), the Law Society's legal aid deserts mapping, the Institute for Fiscal Studies, the House of Commons Library, the Ministry of Justice's Post-Implementation Review of LASPO (2019) and legal aid statistics, the Legal Aid Agency, the Bach Commission's "The Right to Justice" (2017), the Council of Europe's CEPEJ reports, the Equality and Human Rights Commission, Amnesty International, the Public Law Project, research by Dr Jo Wilding (University of Sussex) and Dr Graham Cookson, and the judgments in IS v Director of Legal Aid Casework and Re Z, among others. Figures are for England and Wales except where identified otherwise. Several distinctions matter for accuracy and are observed in the text: the "almost a million to just over 100,000" figure refers specifically to legal help / early-advice matter starts, while the 59 per cent figure refers to completed civil cases between 2013 and 2016; the £728 million / 28 per cent figure is for total legal aid spending, criminal and civil combined, with civil-specific cash falls being smaller; the £350 million was a projected saving for 2014-15, against which the NAO later estimated actual reductions of around £300 million a year; ECF grant rates improved markedly after 2014-15 litigation, and the roughly 1 per cent figure is specifically the first-year position; some downstream-cost estimates derive from advocacy sources and are flagged as estimates, consistent with the NAO's own finding that robust government data on wider costs does not exist; and CEPEJ per-inhabitant figures bundle criminal and civil legal aid, change definition across editions, and are treated as directional rather than precise. The "false economy" charge is advanced here in its defensible form - that the government never established whether the saving survived its downstream costs, not that the costs are proven to have exceeded the saving (Cookson's own estimate implies a net saving roughly half eroded). The most recent and load-bearing figures have been checked against current sources: the September 2025 reshuffle (Mahmood to the Home Office, Lammy as the tenth Lord Chancellor in a decade - or, counting from 2010 and adding Kenneth Clarke and Chris Grayling, the twelfth person to lead the MoJ) and the Civil Legal Aid (Procedure and Remuneration) (Amendment) Regulations 2025, which came into force on 22 December 2025 and delivered effective Controlled Work uplifts of around 42 per cent for housing and 31 per cent for immigration and asylum, not applying to matters opened before that date.