The main concern that current British citizens have about accepting future citizens is that they should integrate into and contribute to society. This is true for people across the political spectrum and every region of the country. While there is disagreement about Britishness and what should be on citizenship tests, a solid consensus favours integration and contribution.

This is summed up well by one person I spoke to, who said that ‘being a British citizen is about being a part of a community’. The key is that the path to becoming British should focus on new citizens coming together with existing citizens. This can be difficult with migrants who come with their own cultural attitudes – they may not choose to adapt or relate them to the attitudes in their new British community.

Integration and contribution can be promoted through a shared language – and successive governments have pursued integration in this way. It is widely popular with the public. Judith from the south-west told me that ‘if you’re going to live here, you should have some basic language skills’. A Russian citizen wanting to become British named Ekaterina went further, saying that a working knowledge of the English language should be the main requirement to be a UK citizen.

Similarly, when Joshua – a student from the south-east – told me that new citizens ‘must be able to integrate’, he made a point of saying that this meant they ‘should be able to communicate in English’. If they could not speak English, Joshua said, this is not good for either them or the community. This was echoed by Sarah, a student living in London, who said that being incapable of speaking English was a problem because it ‘can be isolating for people to be unable to integrate’.

The public are more united about the need for new citizens to integrate into their new society than they are about whether immigration is a problem. This also chimes well with the government’s position. Former communities secretary Eric Pickles said: ‘Speaking English is crucial for everyone to get on, integrate in their communities, interact with their neighbours and gain employment, and our society is weaker when people cannot speak the language.’ For once, we all seem to be in agreement on an immigration policy.

The public’s understanding of integration is based not merely on an acceptance of British values – however ethereal and contested these may be – but on an ability to share in conversations. For most people I spoke to, the level of English they thought mandatory was a basic understanding sufficient to engage with others in everyday British life. New citizens should be able to listen and respond, but also have some written skills.

Some people went further and told me – like John, a British-born citizen from the north-west – that ‘we should develop an oral exam in which we assess the values of the applicant focused on tolerance, humility and togetherness’. An oral exam would help guarantee that new citizens could communicate effectively – whereas an online multiple-choice test could be passed through luck. This is a minority view, but it illustrates the kind of importance most people give to knowing English.

Knowing English has been a central requirement for new citizens for over 100 years. Long a part of the common law, it was incorporated into statute through the British Nationality and Status of Aliens Act in 1914. This is one area where the law seems to have got things right.

But there is also a twist. The Act permitted certain exceptions, which also had deep roots. New citizens were required to satisfy the Home Office they had sufficient knowledge of English – or a language with ‘equality’. This has included Scots Gaelic and Welsh, but not Ulster Scots or Irish Gaelic. As a consequence, a Welsh speaker from Patagonia migrating to the UK would be exempt from demonstrating any knowledge of English.

The legal equality of English, Scots and Welsh continued with the introduction of the Life in the UK citizenship test. While the test handbook has only ever been published in English, it used to be possible to sit the test in Scots or Welsh. This has happened only once: it was sat in Scots in 2010, and was never taken in Welsh since the test was launched in 2005. These facts came to light thanks to a Parliamentary Question asked by native Welsh speaker Roger Roberts, a Liberal Democrat peer from Llandudno, who is the vice-chair of the All Party Parliamentary Group on Migration and a leading champion of migration issues in Parliament.

Perhaps surprisingly, many people recognise that exceptions to knowing English should be permissible. What seems essential for them is that migrants can effectively communicate in everyday life with people in their local communities. If this is an area in Scotland where many speak Scots Gaelic, then this language would be a fitting substitute for English. Jonathan, a Brit born in Scotland, said that ‘you can be a productive member of society without knowing English’. He made clear that ‘not everyone can learn English anyway – personal circumstances must be taken into account’. Many would agree with him – and not only people living in Scotland, either.

This brings us to what I call the English question: how important is knowing English to becoming British?

It might first appear that it only matters insofar as it promotes integration. The great majority of the UK speaks English as a first language, and the importance of English flows from its being the dominant form of communication across most parts of the country. Making exceptions, as in the Scots Gaelic example above, also then makes good sense. This has been a part of the British legal system for a long time and perhaps rightly so.

However, my interviews with British citizens revealed something. No matter which part of the country I surveyed, every region – England, Scotland, Wales and Northern Ireland – put knowing English at the top of the list for what we should expect from new citizens. The Welsh in Wales or the Scots in Scotland did not rank Welsh or Scots Gaelic equal in importance to knowing English.

This finding surprised me somewhat. I expected to find at least a nominal commitment to language equality that was not there. I also expected a greater appreciation for giving other languages equality with English outside England, but this is not the feedback I received. On the whole, nation by nation, English rules.

This may sound like a minor discovery, but it helps explain a massive symbolic change by the coalition government that seems to have avoided any notice.

The equality of Welsh and Scots Gaelic with English, which has existed for decades, if not centuries, is under threat. In 2013, the government changed the rules. Anyone applying for permanent residency or citizenship was now required to know English – without an exception for Welsh or Scots Gaelic. Almost overnight, a longstanding legal right was torn up and no one – not even the Scottish National Party – seems aware of it.

An immediate consequence is that the right to sit the citizenship test in Welsh or Scots Gaelic came to an end. I have spoken to more than one parliamentarian who only discovered this when I told them.

To be fair, the government did make an announcement about the change. But the statement is perhaps more interesting for what it omits than for what it says. On behalf of the government, Lord Bates said:

The ability to speak to and understand the wider community is key to integration. As English is the language spoken throughout the UK, an ability to speak English enables interaction wherever an individual chooses to live. Therefore, a person can only meet this part of the knowledge of language and life requirement by demonstrating speaking and listening proficiency in English.

English should be promoted to best enable integration – perhaps more a government mantra than a fact, but undoubtedly widely accepted. But what this announcement glosses over is the removal of age-old exemptions for Welsh and Scots Gaelic speakers – exemptions which recognised that, for some communities, knowing English may not be enough to foster integration.

It might be said that English is sufficiently embedded into the lifeblood of the UK that any Welsh or Scots Gaelic speaker will surely possess fluency in English as well.

This is probably correct. It is certainly true that part of the reason only one person has sat the citizenship test in a language other than English is because the test handbooks are only available in English. (Unfortunately, there is no public information about whether the person who took the test in Scots Gaelic passed.) This might be seen as evidence that the equality of English with Welsh and Scots Gaelic has lived on only in theory and not in practice since the test’s launch in 2005.

But, to be fair, there is some confusion about whether the government knows what it is doing. In the first pages of the current citizenship test handbook, there is a howler: ‘The Life in the UK test is usually taken in English, although special arrangements can be made if you wish to take it in Welsh or Scottish Gaelic’ – yes, arrangements the government chose to scrap one month after the new test was launched. It makes one wonder how much of this was planned or well known by ministers and their advisors overseeing immigration matters. (Not to mention, of course, that phrase about the test being ‘usually taken in English’ – a slight understatement, given that only one of over 1.5 million tests was in any other language.)

There is perhaps further confusion in that the oath or affirmation of allegiance and the pledge to the Queen and the United Kingdom required from newly fledged Brits at their citizenship ceremonies can be affirmed in either English or Welsh (if in Wales). But it cannot be said in Scots Gaelic.

Besides this, there is another anomaly. Newly elected MPs are required by law to take an oath of allegiance to the Crown on taking their seats. The wording is prescribed in the Promissory Oaths Act of 1868, which permits the oath to be said in English, Welsh, Scots Gaelic or Cornish. This raises questions about why Parliament allows some oaths of allegiance to the Crown in English only, but in multiple British languages for others. Then again, Britain’s ancient customs are sometimes best grasped by not asking why it is, but for how long has it been that way – although this is still unsatisfactory.

The language requirement was not subject to any formal certification by a test until the New Labour years. Previously, sufficient language ability was assumed unless there was evidence to the contrary. This changed with the Nationality, Immigration and Asylum Act 2002, which required new citizens to take one of two routes to prove satisfactory knowledge of English.

The first was to pass the Life in the UK citizenship test, which was designed not only to confirm the applicant’s knowledge of everyday British life and values, but also to guarantee they had satisfactory English – so that people with poorer English will fail. As a consequence, it was written in a terrible style. An Australian who sat the test told me she was surprised to find that a paper that ‘lacked plain English clarity’ was supposed to somehow confirm English fluency. Others said the test was a model in how not to write sentences or ask questions – the awkward formulations were thought to be designed to mislead even native speakers and encourage them to make mistakes.

The second route was via the ESOL with Citizenship course. ESOL – an acronym pronounced ‘Ee-sol’, which stands for ‘English for Speakers of Other Languages’ – is a course of study with many benefits and costs. There is no equivalent in Welsh or Scots Gaelic acceptable as an alternative. ESOL with Citizenship was designed for people whose English is generally poor, opening up a path to citizenship for many people that might otherwise be closed.

Unsurprisingly, this pathway was popular. From 2005 to 2013, more people took this option than the citizenship test. However, solid figures are difficult to come by. This may be explained by the diversity of ESOL provision across multiple providers delivering a suite of different programmes.

An estimated two million people have sat ESOL programmes to date. From what is known from available data, approximately 1.07 million people passed the programme between 2005 and 2012 for a pass rate of 75 per cent, comparable to the citizenship test results. But these figures must be treated with some caution as all are suspiciously rounded up to numbers ending in 00.

There are several downsides – and the ESOL with Citizenship pathway to citizenship has been closed since 2013. It assessed listening and speaking, but not writing. What’s more, people learned about British citizenship as they were being introduced to English; language classes were taught using information found in the citizenship test handbook. This led to some odd-looking assessments, such as the first sample quiz I read, in which people without any prior knowledge of English were asked about the Houses of Parliament. This was probably not what most citizens regard as the first priority for learning a new language. It was certainly a far cry from my first German lessons, where we began with greetings and progressed to reading menus or learning to ask for directions. If learning English is important for integration purposes, something may be missing in ESOL’s approach.

A bigger worry was finding the right class. While the citizenship test was managed by a single organisation, there were a multitude of ESOL and ESOL with Citizenship providers. This means that, unlike the set fee charged to sit the citizenship test anywhere in the country, the costs could vary considerably depending on where it was taken and what baseline level of English the student was starting from. A cursory glance at prices reveals a range from at least £200 to as much as £800 – all a lot more than the £50 citizenship test.

It can be difficult to get a place in an ESOL or ESOL with Citizenship class. Waiting lists vary as widely as their costs. Some places can take on new students on short notice, but many have waiting lists that can stretch for several months. It can take people with little English living in areas of high ESOL demand like London or Birmingham a year or more to achieve, from what I have heard from ESOL teachers. A postcode lottery appears to exist.

This might be a significant issue when it may take someone several weeks to a few months to pass and so continue with an application to the Home Office. It can easily take six months from contacting a college or educational institution for a place to completing a programme of study. If they leave it too late, an applicant may be at risk of overstaying on their visa and facing deportation. It goes without saying that the Home Office does not send emails warning prospective citizens in advance that it may take extra time to get past their red tape. If they can fail an applicant, they often do.

ESOL with Citizenship was a unique programme that some might find a weird way to learn English, but it was wonderful too. An educational manager who taught the course in Leeds told me it ‘can build solidarity for students to help welcome and integrate them while they study’. The classes gave people a reason to come together. They were not just learning about English skills or British values, but learning from each other as migrants in Britain. Building lasting relationships and real friendships with others were some of the most positive aspects of the programme, and were treasured by participants – a perhaps unintended, but highly prized result.

What this indicates is that if English language is important as a means to integrating, we should consider not only the skills or words migrants learn but how they learn them. Working together in a group of people in a similar position can have clear benefits. The advisory group led by Bernard Crick that set out how the citizenship test should work had also recommended that everyone making an application to become British should sit citizenship classes together. One reason was to ensure information was learned and not retained purely for the test. But a second reason was that it was seen as important to bring migrants together so they could share experiences and help support each other’s integration into society. These early plans for a programme sat by all prospective citizens eventually became a course for only those taking ESOL with Citizenship. Better some than none.

Now that this pathway to citizenship has been closed, all applicants have to both pass the citizenship test and provide evidence that they possess sufficient English skills, unless exempted. ESOL courses remain – and the good news is their focus is on English, not on exploring parliamentary democracy while trying to learn how to ask for a cup of tea at the same time.

The programme’s closure was not about making it easier to pass this requirement – it made it much more difficult for people to pass. The reason behind this change is simple. The citizenship test is cheaper and takes less time to do, but it required people have an intermediate level of English ability. That is not a problem for anyone able to read this book (I hope!), but it is for someone illiterate in their first and maybe second languages before trying to learn English as a third.

The ESOL with Citizenship was a helpful alternative. People would be checked to see what level of English they had. To gain a pass certificate, they would have to improve by one level. This meant that someone coming to study English with virtually no knowledge of the language must improve their skills – but need not reach anything like a GCSE level to pass.

Now, the ESOL with Citizenship is no longer an alternative to the citizenship test for those seeking naturalisation, as the test must be sat by everyone – and ESOL courses have become even more popular. They can often include some of the content from the earlier programme incorporating citizenship. I was told that the examples that were discussed as part of ESOL with Citizenship have been useful in ESOL programmes since, to get students talking, in English, about their experiences as migrants and so develop their speaking and listening skills.

Everyone is required to reach an ‘intermediate level’ – which is conversational English. This is easier for some to achieve than others, depending on overall literacy skills generally and baseline level of English specifically. It is obvious that the less English someone knows at the start, the more ground they must cover before they qualify – requiring far more time at far greater cost.

At a workshop run by the National Institute of Adult Continuing Education (NIACE), some found that it could take a beginner up to three years to reach the current standard. A private survey of ESOL providers found waiting lists at up to 80 per cent of the reported institutions. The numbers of people waiting ranged from ten to 1,000, but were typically between 200 and 400.The main problem identified was insufficient provision for low-level learners, especially those with basic needs.

The majority of people taking ESOL courses are migrants on valid visas, including successful asylum seekers. Many classes are chronically underfunded. Despite the government’s belief that such courses help promote much-needed integration – a position endorsed by David Cameron himself in July 2015, when he said that we must ‘ensure people learn English’ as part of his government’s counterterrorism policy to tackle extremism – there were further cuts planned of up to £45 million. This is on top of a steady decline: ESOL funding already dropped from £212 million in 2008/09 to £128 million in 2012/13, so investment is only slipping further.

This is a serious concern for providers of English language provisions, where – according to Dr Melanie Cooke, a leading expert in migrant language education based at King’s College London – ‘everything is driven by funding’ and ‘funding cuts are a major problem’. Not only is it becoming more expensive to teach migrants English, but as funding shrinks, the costs rise even more because more students need to progress much further than before.

In addition, an often overlooked problem is a lack of computer literacy, which further complicates learning English, particularly given how digitally driven much of the delivery can be. Many may run into problems when sitting the citizenship test on a computer if lacking familiarity with how these machines work. Matthew Henderson, an ESOL teacher in West Yorkshire, told me there can be little support for people with learning difficulties or physical disabilities affecting their vision or hearing, raising concerns about equality. Philippa Scutt, an ESOL programme manager based in the Midlands, said there were additional challenges in teaching women from some cultural groups who were unable to use public transport and would only be taught by other women. This required going to them – which she and many others like her do – thus making teaching even more costly.

Cuts to state funding of ESOL programmes have been coupled with a reduction of approved English test providers to only two from 2015. James Brokenshire, the immigration minister, said this is part of their effort at putting in place ‘new, strengthened arrangements’ to end ‘significant abuses’ that have been uncovered.

Saiqa Bibi and Saffana Abdulla Mohammed Ali challenged the Home Office’s requirement that they produce a certificate confirming satisfactory knowledge of English prior to entering the UK in the courts in 2014. This was another effort at raising the standards of English that migrants needed to come to Britain. Each woman was the spouse of a British citizen. Their cases were unsuccessful in the courts. Lord Justice Maurice Kay ruled the government’s test was relatively simple and achievable within a short time frame. Given its purpose was to help facilitate integration of non-English-speaking spouses, the policy was proportionate, if only because the standard required was deemed ‘rudimentary’. This might still be a significant challenge for another who is illiterate in their native tongue.

ESOL is not required for everyone. While tightening the regulations, the Tory-led coalition government also introduced a wide range of exemptions – nearly a dozen in total. Some of these are sensibly based on age, and include exemptions for children and persons over sixty-five years old.

Others are based on nationality. These are a problem because they expose a major loophole in the policy. Several countries listed for exemptions are those we might expect. Citizens of Australia, Canada, New Zealand and the United States need not demonstrate they know a single word of English – they satisfy the criteria by having the right passport. But some other countries that are similarly exempted may be surprising, such as Belize and the Dominican Republic.

The nationality exemptions lack a clear rationale for how countries are selected for inclusion on the list. Countries where English is the official de jure language are sometimes included and sometimes not; those where English is the unofficial de facto language are treated equally erratically. Countries with very high English fluency rates like Singapore – to choose just one example – fail to make the list without rhyme or reason. No guidance has been made readily available to explain why over a dozen countries made the list but others did not.

Additionally, the exemptions confirm an expectation that every national from a listed country will have satisfactory English language skills. This is clearly misguided. There are about 45 million people in the United States for whom English is neither a first nor a second language. Many more Americans lack the intermediate standard of English that the British government wrongly assumes they possess.

Exemptions are also made for anyone with a degree taught in English – even if it is not from an English-speaking country. To qualify, a person must have an academic qualification recognised by UK NARIC as equivalent to at least a UK bachelor degree. NARIC – the UK’s National Recognition Information Centre – is a government agency that recognises and compares international qualifications and skills. This comes with a fee of about £50. Kwame, a Ghanaian whose naturalisation application is currently being processed, holds a degree taught in English, and paid NARIC for a certificate. He told me it took two weeks before it arrived in the post. Frustratingly, he was never asked for it when interviewed for his permanent residency. Kwame said that if he had not had a certificate, ‘the Home Office would not know’. This is evidence that exemptions to the English language requirements are not being enforced as they should.

If qualifications are not checked when applications are processed, it only reinforces the view of most migrants becoming British that many of the requirements serve little purpose other than soaking migrants for as much of their cash as possible – a view that can harden over time and undermine the integration that many of these policies are intended to promote.

So, what records are kept? It would appear none at all. Lord Roberts of Llandudno asked the government to confirm how many people had been exempted from the English language requirement on the grounds that they held a UK NARIC certificate. He discovered that the government is not counting and there are no statistics being collected. They have no idea how many people may be using this route to slip through the net and ‘prove’ a fluency in English they do not possess.

Many might think that requiring a degree such as a BA, an LLB, an MBA or a PhD that is undertaken in English will mean that the English language skills of the graduate will be satisfactory. While this may be true in many cases, it will not be true for all. In 2015, I was asked to review applications from several senior academics for a national funding agency in Kazakhstan, where English is spoken by only a very small group of people. The applications were in English and came from people in charge of awarding degrees taught in English. Take my word for it – their English was appalling. Missing words, spelling errors – you name it, they did it.

Just because English language instruction may be on the menu at a university or college somewhere in the world does not mean its students will all be OK. It is very possible that many lacking adequate skills will slip through the net. Degrees are no doubt beneficial for finding work, but they are no guarantee the student’s English is up to scratch.

This is an important issue, as government ministers may not know exactly how good the quality is of English certificates from non-British universities. The immigration minister James Brokenshire even admitted, in response to a Parliamentary Question from his Conservative Party colleague Graham Brady MP, that the Home Office lacks data on the number of cases of fraud or abuse in English language testing centres abroad. This answer should satisfy no one.

So is this really about integration or economics? Phil Woolas, a former immigration minister, told me that government is always looking at the issue of ‘does it benefit Britain?’ He said that this view is short-sighted because we must account for how local Britons perceive social change and the impact British immigration policy may have on the countries of origin for the people who come to the UK. Phil is right.

Providing more accessible ESOL training for migrants, boosting their employability skills, helps them better compete against native Brits – potentially stoking misgivings about the value of immigration for the UK. Those struggling to secure employment may feel even more left behind. We need to think more carefully about the consequences, both intended and unintended.

It is clear that much more work needs to be done. According to the last census, over 850,000 migrants reported that they could not speak English well or at all. The Times revealed in 2014 that a quarter of councils in England and Wales have areas where little English is spoken. They found from the last census in 2011 that 138,000 people in these areas could not speak any English while nearly 750,000 could not speak it well. If being able to converse in English is so important for integration and making a contribution, there are too many people left out who need to be brought in.

One ESOL teacher told me that often his students would be in the country for several years and only come to him when they realised they had to fulfil this requirement. The longer they waited, the more difficult learning became, as the deadline for their expiring visa could be heard virtually ticking in the background. Studies also suggest that the longer someone waits before starting English classes, the slower their progress. A welcome change would be to help encourage people to get started much earlier. But this is easier said than done.

There are diverse ways of learning English and no set national curriculum for getting from zero to hero. This flexibility to best suit the particular needs of individual learners is welcome and should remain. But many teaching on ESOL programmes complained that the quality of study materials for the citizenship test was much better than the quality of those for the English language tests.

This shortfall should be addressed and government must look into making it easier and more rewarding to learn English. It is in the national interest and also in the interests of every one of us living in Britain. Making ESOL compulsory for more people while providing less funding does not help. The government must consider whether it is a price worth paying to make it more difficult to learn adequate English skills, given its importance for integration and social cohesion and given their efforts at countering extremism.

The government should take the importance of English more seriously. It can begin by ending unnecessary loopholes. Exemptions for the very young and for pensioners should remain, but exemptions based on nationality should go. A recommendation I made to Ed Miliband’s home affairs team in regular meetings prior to the 2015 general election manifesto – but which was not taken up – is that there should be an expectation that everyone pass an English test. No exemptions based on nationality or holding a degree. Perhaps it will be very easy for some to pass. But that is not the point. It is in the public interest that new citizens can converse in English, and tightening the requirements much more than the current government is doing would be a welcome step in the right direction. Keeping exemptions to a minimum would help streamline the process and shrink the associated unnecessary bureaucracy.

There might be a fee of £50 to take the listening and speaking English test. This new fee could reduce expenses for applicants and cost less than UK NARIC certificates for degree-holders. People I spoke with who deliver ESOL programmes thought the influx of additional revenue would help make their programmes more sustainable by creating a new source of income. This would come from migrants and not burden the taxpayer. The extra funding created could go to improving provision for the lower-ability learners where it is needed most. And this would not increase the costs for those who might otherwise be exempt through paying UK NARIC for a certificate – everybody wins.

Lara, an Italian national in London trying to become British, told me, ‘What has to be done to prove English language ability was silly’ – especially when anyone could simply speak with her and see she is fluent. There should be a simple test for her and everyone else that holds all to the same standard. Fluent speakers like Lara could directly sit the required brief test, saving time and money. More importantly, we can become more certain than we are now that new citizens really are suitably fluent in English.

There are English language test providers that conduct assessments for the UK Visas and Immigration department of the Home Office, valid for citizenship applications already in place. The assessment consists of a ten-minute speaking and listening test, with results normally available that day and a certificate received in the post within seven days. The costs can be as high as £150. This could be reduced if more took it as a standalone assessment with a standard test and fee – as there is with the citizenship test. Migrants like Lara do not need a several-week language course and should not be required to take one. But a fair test that all are expected to pass would better ensure standards are maintained, produce much-needed income for ESOL providers and best secure policy aims.

Taking English provision seriously should include a wider approach. The United States runs their test provision very differently. It covers writing in addition to speaking and listening, but it also provides a wide array of free materials to help prospective new citizens learn English. This includes vocabulary flash cards, self-tests and downloadable videos. Likewise, the British government should adopt a similar high standard and provide more materials for free online. This would better promote raising the bar on literacy for all – it is not only migrants who should have good English, but everyone.

Literacy is not a problem for migrants alone. Illiteracy rates in the UK are surprisingly high. While only 1 per cent of the UK is fully illiterate, there are roughly 16 per cent who are ‘functionally illiterate’ and have the literacy levels of an eleven-year-old or worse. That’s over five million people.

If the government is so committed to fluency in English to boost social cohesion and employability, then requiring literacy from immigrants addresses only one part of the problem. If literacy is good for all, it should not be required from only some. There is a case for providing free English instruction for all. Let migrants pay a fee like £50 for certification if planning to apply for citizenship, but the availability of English support should not be aimed primarily at migrants alone when so many native Britons remain functionally illiterate too.

The government recently changed part of its strategy. In January 2016, David Cameron made a statement that improving English standards was central to ending segregation in the UK and tackling extremism. He promised £20 million to support ESOL providers to deliver English training to women-only groups, as women lacking English fluency were deemed to have greater needs than men. Cameron also announced new English language checks for everyone on spousal visas after two and a half years – or halfway to the five years of residency in the UK required for permanent settlement – to ensure progress is being made.

These measures are not as helpful as they might look. The £20 million is presented as ‘new’ funding, but it is much less than the £45 million slashed from ESOL funding in 2011 and 2015, during Cameron’s time in 10 Downing Street. While funding was available to support men and women before, less than half has been restored to help women alone. Although English training for women in single-sex groups is crucial for engaging with some communities, and delivering more of it is a welcome development, this support has already been provided for some time by many existing ESOL programmes.

Cameron’s plan to introduce mid-visa English progress checks sounds like a tougher standard, but that is an illusion. Under New Labour, spouses had been able to apply for permanent residency, ahead of a citizenship application, after two years. This meant that anyone wanting to continue living permanently in the UK needed to meet the required English standard in two years. But Cameron lengthened the residency requirement for spouses from two to five years. This meant that spouses had more time to reach sufficient English fluency and they could delay their learning. It is doubtful that any mid-visa review is practical or the consequences of failing it would be lawful. There would need to be clarity on what qualifies as ‘mid-visa’, which could become unnecessarily complex – and it is difficult to see the legality of rejecting settlement by a fluent English speaker who started learning more slowly but had reached the required standard by the time they applied. In short, the government’s changing strategy is more a bag of gimmicks that steady progress.

Other European countries take a harder line on language and integration that we might learn from. Germany offers integration classes (Integrationskurse). These teach migrants German language, history, culture and basics about its legal system. Failure to attend can lead to cuts in eligible benefits. France has a similar system based around ‘integration and welcome contracts’ (contracts d’accueil et d’integration), where attendance at classes is required and monitored. These programmes are more to the spirit of Bernard Crick’s recommendations a decade ago, which have yet to be put in place, where migrants are to come together to learn, share and transition to citizenship.

The German and French models are not perfect. I understand that migrants do not receive much support in learning about and joining integration programmes. Applications for citizenship in France are more likely to be rejected than accepted than in the UK. This suggests that either such integration programmes should do a better job at supporting prospective citizens to become naturalised or, if the programmes are well run, that more should have access to them. My point is that if integration has crucial significance and language ability is a key means to integration, then we should look to establish more opportunities for future long-term residents and citizens to benefit from integration programmes. They exist elsewhere and should be here, too.

Perverse incentives exist in the UK system that undermine the way it works. For example, someone on Jobseeker’s Allowance can receive funding to help cover the costs of completing an ESOL programme – financial support that might be lost if he or she obtained work while taking ESOL. The problem is not only the disincentive to choose work where possible, but the priority of obtaining the certification over acquiring skills.

If it were about improving skills, more focus would be placed on improving English over the long term and not merely until it is good enough for immigration purposes. For example, Home Office guidance states that some English language qualifications on its approval list expire as qualifications after two years. However, they ‘will still be accepted as proof if you’re applying for citizenship, even if it has expired’. It does not matter how many years ago the qualification was earned or expired – if you have it, you are OK. But this reduces the importance of knowing English to a mere box-ticking exercise for little clear public benefit.

Finally, ESOL teachers made clear to me that what they wanted most of all was for English language teaching to become depoliticised and centred on an effective, long-term strategy. In a report for Demos, Ally Paget and Neil Stevenson diagnosed problems with current ESOL policy very well. They said it ‘suffers from fragmentation, a lack of clarity about the aims and intended outcomes of learning, disagreement over the analysis and description of English language levels and abilities, and a general tendency to take a short-term view’. While funding has moved to a one-size-fits-all model, the range of learners becomes more diverse as standards are raised and many more must pass to stay. Such a situation can only stoke frustration and anguish for those who engage with the system – and may do little to help promote the integration that is supposed to drive the policy in the first place.

Producing a long-term strategy might also help to tackle the many blurred lines of accountability. For example, the Welsh government is responsible for managing the provision of ESOL as a part of Lifelong Learning. However, it is the British government that has responsibility for immigration and citizenship policy for everyone, and this can affect the provision available in Wales. There are additional challenges in learning English in Wales, not least recognising and understanding Welsh place names and signs ‘when English is “alien” as well’, as the Welsh government acknowledges. Requiring English can mean supporting different approaches in different parts of the UK.

There is an eye-wateringly long list of bodies that inform ESOL training and delivery, including the Home Office; the Department for Business, Innovation and Skills; Ofqual; the Skills Funding Agency; the Department for Communities and Local Government; and the Department for Work and Pensions (through Jobcentre Plus), each of which feeds into ESOL policy or provision one way or another. A streamlined, joined-up structure could bring clarity and improved accountability.

No government can guarantee every citizen knows English. There will be children born abroad to British parents who may never hear a word of it, but can still earn a British passport – this is not a problem. But the government can do much more than it currently does to ensure more British citizens are literate – whether they are born British or become British. The question is not whether the necessary reforms are beyond our reach – they are not – but rather whether any government will show the commitment to complete the modernisation project launched under Blair and New Labour and win the war against illiteracy and functional illiteracy.