Counterfactual history is easy to abuse.
Give an imagined government perfect foresight, unlimited money and no opposition, and almost any outcome becomes possible. That is not analysis. It is a wish list.
A useful counterfactual is harder. It keeps the pressures of the period: globalisation, deindustrialisation, the expansion of higher education, the 2008 financial crisis, the arrival of smartphones, the demand for fiscal restraint and the 2016 referendum. It accepts that governments would still make mistakes and that no policy could manufacture friendship or prevent every form of loneliness.
It changes something narrower: what government noticed, what it measured and what it refused to treat as somebody else’s problem.
The alternative examined here is therefore not a return to 1990. Britain still modernises. Shops move online. Migration continues. Education expands. Recruitment becomes digital. Smartphones arrive. The question is whether the social functions displaced by those changes are recognised early enough to be preserved in safer, fairer forms.
Counterfactual rule
If a reform removes a common route through which young people encounter work, adults, community or independence, government must ask what will perform that function next.
That rule would not guarantee a less isolated country by 2020. It would give Britain a better chance of remaining socially open while it changed.
1990–1996: preserve the first rung
The first task is not to preserve declining industries or romanticise casual teenage work. It is to notice that entry-level jobs perform two functions at once.
They produce goods and services. They also introduce inexperienced people to the adult world.
Britain did create a promising instrument. Modern Apprenticeships were announced in 1993, piloted from 1994 and expanded across England from 1995. Apprentices had employed status, a wage and a structured qualification. National Traineeships followed as a route for young people not yet ready for Level 3.[1]
The credible alternative begins there. It makes the new system a central institution rather than one programme among many.
Public support would cover part of the real cost of taking a first-time worker: supervision, administration, training and the lower productivity that comes with inexperience. Small employers would receive practical help rather than another portal to navigate. Quality standards, wage protections and safeguarding would remain intact. The answer to employer reluctance would not be weaker protection for 15- to 18-year-olds; it would be help meeting the cost of employing them well.
Alongside apprenticeships, a Teenage First Step offer would support properly supervised part-time work for young people still in education. Employers would be rewarded for providing genuine responsibility and contact with customers or mixed-age teams, not for relabelling cheap labour as training.
Transport belongs in the same policy. A vacancy that cannot be reached at 7 a.m., after college or from a rural town is not a practical opportunity. Youth concessions, evening services and links between schools, colleges and employment sites would be treated as part of the programme rather than left to chance.
None of this would stop the long decline in traditional teenage work. Technology, employer practice and the changing economy would still reduce many entry-level roles. The purpose would be to prevent the first rung from disappearing faster than a credible replacement could be built.
1997–2003: use prosperity to build resilience
The next government still introduces the minimum wage, strengthens employment protection, expands university participation and uses the New Deal to address unemployment. The alternative is not lower standards or fewer academic opportunities.
It is route balance.
Every young person between 14 and 18 would receive repeated, meaningful contact with workplaces. Not one poorly organised week photocopying in an office, and not an expectation that every teenager already knows the right adult to ask. Schools, colleges, employers and local intermediaries would organise a sequence: workplace visits, short projects, supervised placements, mentoring and paid work where appropriate.
The standard would be exposure to the real social environment of work: customers, deadlines, unfamiliar adults, minor conflict and responsibility. A placement with no meaningful task would not count.
Government would also begin measuring the things later lost between departmental silos: youth employment while studying, workplace exposure, independent travel, participation in clubs and volunteering, and time spent outside the home. The measures would be published by age, income and geography. They would not yet be rolled into a grand index. The first objective would be to see the pattern clearly.
This matters because prosperous years are when preventative infrastructure is easiest to build. Waiting until youth unemployment rises or local services close makes every response more expensive.
2004: manage enlargement without inventing a villain
EU enlargement still happens. Britain remains open to migration and recognises its economic and human benefits.
But it uses the temporary labour-market restrictions available under the accession arrangements, as most existing member states did. The restrictions would be time-limited, reviewed openly and linked to labour-market evidence rather than treated as a permanent barrier.[2]
The purpose would not be to reserve jobs by nationality. The evidence does not establish that migration caused the long decline in British teenage employment, which had already begun. A credible government would say so.
The narrower concern is coordination. A rapid increase in the supply of experienced adult labour could affect employer incentives in sectors that also supplied first jobs. At the same time, policy was directing more British teenagers towards education. Those movements should have been examined together.
During the transition, sectors reporting shortages could open earlier where the evidence justified it. Large employers in retail, hospitality, logistics, care and food production would face the same training expectation regardless of whom they hired: publish entry-level recruitment, invest in domestic training and offer credible progression from first job to skilled work.
Making the duty nationality-neutral matters. Conditioning support on the use of migrant labour would risk discrimination and confuse the issue. The failure to train is an employer and sector problem, not a trait attached to the worker who fills the vacancy.
There would be costs. Some employers would face tighter labour supply; growth in certain sectors might be slower; transitional controls could encourage informal work. Those risks would need monitoring. The counterfactual claim is not that restrictions were costless. It is that Britain had a lawful opportunity to slow the transition while building institutions it already knew were weak.
2007–2012: participation without confinement
The case for raising participation after 16 remains strong. Leaving education with few qualifications carried serious long-term risks.
The alternative rejects only the idea that participation must mean remaining primarily inside an educational institution.
From 16 to 18, recognised routes would include academic education, a high-quality apprenticeship, paid employment with accredited training, a substantial vocational programme or structured civic service. These routes would not be declared magically equal by law. They would be made credible through comparable funding, clear standards, progression into further study or work, and an entitlement to impartial guidance.
Work-related learning at Key Stage 4 would be retained but rebuilt. The Wolf Review was right to challenge low-value placements. The answer would be quality assurance, not removal of the bridge. Schools would be judged on whether students encountered real workplaces and understood several routes into adulthood, not on whether a box had been ticked.[3]
Education would still protect young people during a weakening labour market. It would no longer be expected to substitute for every form of adult contact.
2008: rescue the first years of working life
The financial crisis still arrives. Banks are still rescued. Public borrowing still rises sharply. The difference is that the social scarring of youth unemployment is treated as part of the emergency, not a secondary problem.
Britain eventually introduced the Young Person’s Guarantee and Future Jobs Fund. By January 2011, official statistics recorded 91,890 starts in Future Jobs Fund vacancies, alongside starts in community placements and training.[4] The counterfactual does not invent a programme with no historical basis. It begins the response earlier, makes it larger during the worst of the downturn and protects its strongest elements for longer.
Anyone under 25 unemployed for four months would receive a credible offer: an apprenticeship, a time-limited subsidised job, a paid community role, training tied to a vacancy or an intensive employment placement. The promise would be an offer, not the impossible claim that nobody could become inactive.
Jobs would have minimum hours, supervision and progression standards. Employers would not be paid merely to replace an existing worker with a subsidised one. Local government, charities and small businesses could propose useful roles, but poor schemes would be closed quickly.
The policy would carry deadweight costs: some funded jobs would have existed anyway. Some placements would fail. That is the price of acting at speed. Transparent evaluation would compare earnings, employment and participation after the subsidy ended. The goal would be to prevent a temporary crash from becoming a permanent absence from work and adult society.
Housing: act before dependence hardens
Housing is the most difficult part of this alternative because no single policy could have made homes abundant and affordable.
Earlier action could still have reduced the damage.
During the credit expansion of the 2000s, regulators would stress-test affordability and monitor high loan-to-income lending before household debt became a systemic concern. The Bank of England eventually imposed a loan-to-income flow limit and stronger interest-rate stress testing in 2014. An earlier macroprudential body could plausibly have acted sooner, although the effect on prices would have been uncertain.[5]
Credit restraint alone would not solve undersupply. It could also make ownership harder for some first-time buyers. It therefore has to sit beside steady housebuilding, planning reform, affordable rental supply and stronger security for tenants.
The social measure would be simple: policy would track the age at which people could establish an independent household and distinguish preference from financial constraint.
Living with family would never be treated as failure. Being unable to leave would be recognised as a restriction on autonomy with consequences beyond the housing market.
This would not produce cheap homes everywhere by 2020. It might reduce speculative pressure, improve rental stability and make delayed independence harder for government to ignore.
2010: protect the places that ask little of you
Fiscal consolidation still occurs after the crisis. The alternative changes its speed and its hierarchy of protection.
Youth services would be classified as preventative infrastructure. England already had a statutory framework: section 507B of the Education Act 1996 required local authorities, so far as reasonably practicable, to secure sufficient educational and recreational leisure-time activities and facilities for young people.[6] The problem was not the total absence of a duty. It was how weakly “sufficient” could be defined, funded and enforced.
National government would set a minimum service expectation while leaving local authorities room to decide how to meet it. Every area would publish what young people could reach, when it was open, what it cost and how provision differed across neighbourhoods. Funding would protect a basic network of youth work, libraries, community sport and inexpensive places to gather.
This would require slower consolidation, reprioritised spending and, in some years, modestly higher borrowing or taxation. It should be stated plainly. By 2016–17, council spending on local youth services had fallen from £650 million in 2010–11 to £390 million, while more than 600 youth centres were reported closed across the UK between 2012 and 2016.[7] Preventing a large part of that contraction would cost real money.
National programmes such as the National Citizen Service could add valuable intensive experiences. They would not be treated as substitutes for a local place open every week. A residential programme and a familiar youth worker perform different functions.
The right to move — and to remain
Young people need places. They also need the freedom to reach them and some confidence that they will be allowed to stay.
The alternative would treat independent mobility as a gradual capability, not a choice between total supervision and neglect. Safe routes to school, lower traffic speeds, youth bus concessions, evening and weekend services, rural transport links and secure cycle routes would make independence more realistic. Schools and councils would help families build age-appropriate freedom in stages.
The risks would not be dismissed. Roads can be dangerous. Harassment and violence are not distributed equally. Girls, disabled young people and minority groups may experience public space differently. Safer design, trusted adults and reliable transport are therefore part of independence, not arguments against it.
Public-space policy would also distinguish behaviour from mere presence. Dispersal powers would remain available for genuine intimidation or disorder, but their use would be recorded and reviewed by age and location. Devices designed simply to repel young people would not be used by public bodies.
The principle is modest: a teenager behaving lawfully should not need to be buying something to have a legitimate place in the town they live in.
2011–2015: study digital displacement before claiming causation
No credible government in 2007 understands the full effect of the smartphone or social media. The alternative does not pretend otherwise.
By the early 2010s, however, it is reasonable to ask a disciplined question: what is digital activity replacing?
Government would fund longitudinal research that separates different activities, ages and users. Messaging a close friend is not the same as passive scrolling. Online support for an isolated child is not the same as compulsive engagement. Correlation between heavy use and distress would not automatically be presented as proof of causation.
Early measures would focus on what was already knowable: strong privacy defaults for minors, limits on unnecessary data collection, accessible reporting, media literacy and independent access for researchers to appropriately protected data. Schools would teach attention, verification and online conduct without pretending abstinence was a realistic national policy.
Only as the design of engagement systems became clearer would regulation move towards persuasive features, profiling and recommendation systems. The sequence matters. A detailed ban on manipulative notifications in 2011 would be anachronistic. Building research, regulatory capacity and privacy protection from 2011 would not be.
By 2018, Britain had a concrete legislative route. The Data Protection Act required the Information Commissioner’s Office to prepare an age-appropriate design code. The first draft went to consultation in 2019 and the code was laid before Parliament in June 2020.[8] In the alternative, this work begins earlier, is better resourced and is accompanied by enforceable defaults for services likely to be used by children.
The crucial difference is that digital regulation is paired with offline provision. Telling a teenager to spend less time online while closing the youth club and cutting the evening bus is not a serious policy.
Recruitment keeps a human exit
Digital recruitment brings advantages. It widens reach, creates records and can reduce some forms of informal favouritism. It also allows rejection to become invisible.
By the late 2010s, employers receiving major public contracts or training subsidies would be required to audit automated screening for adverse effects and tell applicants when automation materially shaped a decision. Young candidates would have access to a workable alternative route, an explanation or proportionate human review.
That is not an interview for everyone. Large employers could not function on that basis. Nor is human judgement automatically fair.
The aim is narrower: an inexperienced applicant should not be excluded by an opaque system with no realistic way to correct bad data, explain an unusual route or demonstrate potential.
Efficiency can remain. Final responsibility cannot disappear into software.
2016: respect the vote, change the bargain
The referendum still happens and its result is respected.
But 2016 does not instantly produce the post-Brexit immigration system; that belongs to the period after Part I. What can change before 2020 is the bargain being designed.
Sectors seeking easier international recruitment would publish evidence of shortage, pay, retention and training. In return for access where it was needed, employers would commit to apprenticeships, first-job recruitment and credible progression. These expectations would apply across the workforce, not cast migrants as the cause of domestic policy failure.
Migration would remain part of Britain’s economy and society. It would no longer serve as an excuse for government and employers to postpone the harder work of building entry routes at home.
2017: make the levy protect beginnings
The apprenticeship levy is retained. Its purpose — forcing sustained employer investment in training — is sound.
Its design would include guardrails from the start. A substantial share would support under-21s, first-time occupational entrants, small employers and Level 2 or Level 3 routes with proven progression. Existing staff could still undertake higher apprenticeships, but levy funds could not drift mainly towards rebadged senior training.
This is not a theoretical concern. Between 2015–16 and 2018–19, apprenticeship starts among under-19s in England fell by 26 per cent. In the first quarter of 2019–20, starts among 16- to 19-year-olds were 21 per cent lower than a year earlier. Over the same period, the share of starts at higher levels rose while the share at Level 2 fell.[9]
Numbers alone would not be the target. Completion, wage progression, occupational skill and whether an apprenticeship created a genuine new entrant would matter. A smaller number of good apprenticeships can be better than a larger number of weak ones. But a training system for all ages should not quietly cease to provide a first step for the young.
2018–2020: make social opportunity visible
By the end of the decade, Britain would not unveil a scientifically complete Social Contact Opportunity Index. The data were too fragmented and the weighting too contestable for that claim.
It could build the machinery honestly.
A public dashboard would track six dimensions: workplace exposure; independent mobility; youth and community infrastructure; physical journeys and affordable third places; economic and residential independence; and direct participation in sport, volunteering and face-to-face activity. Digital connection would appear beside them, not be counted automatically as an equivalent substitute.
Departments would pilot a short Social Connection Impact Note for policies likely to change daily behaviour. It would ask:
- Does this remove a routine reason to enter shared physical space?
- Does it make independent movement easier or harder?
- Does it change access to work, trusted adults or mixed-age environments?
- Who can purchase a private replacement, and who cannot?
- What evidence will show whether the effect occurred?
Most policies would need only a brief assessment. Major reforms to transport, youth provision, education, housing, planning or public-service digitisation would require more. Pilots would test whether the process revealed anything useful before it became another bureaucratic ritual.
The purpose would not be to maximise contact. People need solitude, privacy and freedom from compulsory sociability. The aim would be to stop government increasing isolation unknowingly because the effect never appeared in any department’s account.
What this path would not solve
This alternative is stronger if its limits are visible.
It would cost money. Youth transport, services, supervision subsidies, housing supply and crisis employment programmes require taxation, borrowing or cuts elsewhere.
It would create friction. Employers would face training duties and recruitment safeguards. Platforms would face design constraints. Planning reform would provoke local opposition. Earlier mortgage restraint could exclude some buyers even as it reduced wider risk.
It could be captured. Subsidised jobs can displace unsubsidised ones. Apprenticeships can become relabelled training. Impact assessments can become paperwork. Statutory minimums can become ceilings.
And it would not control culture. Young people might still prefer online life. Parents might still supervise more closely. Communities might decline to use the spaces provided. The 2008 crash would still damage confidence and earnings. Social media would still bring benefits as well as harms.
The case for the alternative is therefore not that every intervention works. It is that government keeps the social consequences inside the decision, tests what it builds and changes course when the evidence fails.
The path Britain could have taken
By early 2020, Britain would still look modern.
Teenagers would still use smartphones. More young people would remain in education. Employers would still recruit online. Migration would still shape the labour market. Retail would still move towards the screen. Housing would still be expensive in many places.
But more young people might also have a first rung into work, repeated contact with adults beyond school, a bus that ran when they needed it, a public place where their presence was legitimate and a local service not dependent on family income. Fewer might reach adulthood having been protected from risk but also excluded from practice.
The difference is not nostalgia. It is institutional memory.
A Saturday job is not only labour. A bus is not only transport. A youth club is not only council expenditure. An apprenticeship is not only training. A home is not only an asset. A shop is not only a transaction.
Each can be a point where a young person meets society.
Britain could have modernised while protecting more of those points. It did not need to preserve the old form every time. It needed to preserve the function — or build a better one.
The alternative path was not inevitable. Neither was the path we took.
Then, in March 2020, even the stronger system imagined here would have faced its severest test.
That is where Part II begins.
Source notes
The sources establish the historical record, chronology and policy mechanisms. The counterfactual choices remain editorial judgments.
- House of Commons Library, Apprenticeships Policy, England prior to 2010, 2015: Open source ↩ Return
- European Commission, Report on the Functioning of the Transitional Arrangements set out in the 2003 Accession Treaty, 2006: Open source ↩ Return
- Department for Education, explanatory memorandum on the removal of the Key Stage 4 work-related-learning duty, 2012: Open source ↩ Return
- Department for Work and Pensions, Young Person’s Guarantee Official Statistics, April 2011: Open source ↩ Return
- Bank of England Financial Policy Committee, record of meetings held 17 and 25 June 2014: Open source ↩ Return
- Education Act 1996, section 507B: Open source ↩ Return
- Local Government Association, Children’s Social Care in England, 2019: Open source ↩ Return
- Information Commissioner’s Office, background to the Children’s Code: Open source ↩ Return
- House of Commons Library, Effectiveness of the apprenticeship levy, February 2020: Open source ↩ Return
This is a historically constrained counterfactual, not a claim that the policies described would certainly have produced lower loneliness or better mental health. The sources establish the policy context, available mechanisms and chronology. The proposed choices and likely effects remain reasoned judgments.