Overview of Home alone: the relationship recession
This episode of The Economist’s podcast examines a sustained rise in singlehood across the developed world — a “relationship recession” — and explores its social, economic and technological consequences. Journalists Jonathan Rosenthal and Su Lin Wong explain why more people (especially younger adults) are remaining unpartnered, how dating practices and gendered economic shifts play into that trend, and what it means for fertility, housing and society. The episode then turns to technological responses to loneliness — from AI companions and humanoid robots to apps — weighing the benefits and risks. A short third segment explains how the tobacco industry has thrived despite a large fall in smokers.
Key discussion points
- Singlehood has risen sharply over decades and has accelerated in the past 5–10 years.
- Pandemic disruption and increased screen time (video games, social media, apps) contributed to reduced real-world social practice and dating.
- Dating apps change mate-selection dynamics by enabling highly specific filters (e.g., height, politics), narrowing pools and altering matching patterns.
- In Asia, two distinct groups are disproportionately single: low-educated men (often rural migrants) and highly educated urban women seeking similarly educated, egalitarian partners.
- Roughly 60–70% of single people (surveyed) say they would prefer to be in a relationship — many feel lonely or discouraged.
- AI companionship is growing rapidly (apps and robots) and is being used as friends, mentors, therapists or romantic partners.
- Tobacco companies have offset volume declines by raising prices to exploit a remaining, price-inelastic base of addicted smokers, keeping profits high.
Notable statistics & data cited
- In the U.S. (2023), 41% of women and 50% of men aged 25–34 were single — roughly double the rate from five decades ago.
- OECD study (2010–2022): in 26 of 30 countries more people were living alone.
- Bumble anecdote: many women list 6 ft as a minimum height, dramatically shrinking the eligible male pool (claim cited: ~85%).
- Character.ai: ~20 million monthly active users.
- Survey (U.S. high school students): 42% said they or a friend had interacted with an AI friend.
- Harvard Business School working paper: short-term chat with AI companions reduced loneliness more effectively than passive online activities in that study.
- MIT/OpenAI research: heavy ChatGPT users reported higher feelings of loneliness (causation not established).
- U.S. adult smokers down by roughly 20 million in the past decade, yet tobacco equities have outperformed the NASDAQ over a 2-year stretch.
Causes and drivers of rising singlehood
- Structural economic change: women’s rising education and labour-market participation expand choice and reduce economic need to marry.
- Cultural and technological shifts:
- Pandemic halted formative dating experiences for young adults.
- Increased time spent online and gaming reduces in-person socialization and meeting opportunities.
- Dating apps create abundant choice and hyper-specific filtering (preferences on height, politics, education), fragmenting the dating market.
- Gendered mismatch: highly educated women often prefer similarly educated, egalitarian partners; lower-educated men face worse prospects, sometimes marrying foreign partners.
- Backlash and identity politics: some men adopt anti-feminist or “incel” stances; in some countries (e.g., South Korea), female-led movements (e.g., no dating/no marriage) and male anger can reinforce polarization.
Social and economic implications
- Fertility: lower coupling rates reduce birthrates, with long-term implications for demographics and pensions.
- Housing: more people living alone increases demand for units and changes housing markets and planning.
- Welfare and public budgets: aging populations and fewer young dependents will shift government spending patterns.
- Social cohesion and politics: rising involuntary singlehood and online subcultures (incels) can feed political instability and toxic social movements.
- Emotional health: a substantial subset who are involuntarily single report loneliness, lower life satisfaction and desire for relationships.
AI companionship — rise, benefits, and risks
What’s happening
- AI companions range from general chatbots (customized prompts on ChatGPT) to dedicated apps with persona-driven characters and humanoid robots (e.g., Daisy, a DexLab robot using GPT-based chat).
- Tech firms are adding features to make chatbots more humanlike; some plan to allow erotic interactions for verified adults.
- Companion use is growing fast and entering hardware (robots, toys) as well as apps.
Potential benefits
- Short-term reductions in loneliness for some users (evidence from small studies).
- Companionship and assistance for older adults, disabled people, or socially isolated individuals.
- Emotional support and stress relief for people unwilling or unable to form human relationships.
Main risks
- Withdrawal from human relationships: users may prefer sycophantic AI that never challenges them, worsening long-term social skills and resilience.
- Unrealistic expectations: AI’s agreeable behavior could distort expectations about human interactions.
- Mental-health harms and legal exposure: lawsuits in the U.S. have been filed in cases involving teenagers and AI interactions linked to self-harm.
- Volatility and dependency: personality or behaviour changes after model updates can traumatize users who have formed attachments.
- Privacy, manipulation, and loss of agency as decisions and emotional labor get outsourced to commercial systems.
Safeguards and regulation
- Companies implementing parental controls, age-gating and content moderation, but these can be circumvented.
- Possibility of future government regulation, though global coordination and enforcement are complex.
- Importance of transparency about data use, model updates and limits of AI companionship.
Tobacco industry spotlight (brief)
- Falling smoking prevalence has left a cohort of more addicted, price-insensitive smokers. Tobacco firms raise prices to offset volume loss and maintain/increase profits.
- Evidence shows tobacco price inflation outpacing general inflation; firms explicitly cite pricing strategies in earnings calls.
- Limits exist: at some point price increases drive quitting, death, or substitution to vaping or black-market products.
Notable quotes & insights
- “Dating is not like riding a bicycle. You don't just do it once and then remember how to do it.” — on pandemic-era lost practice for younger cohorts.
- App-era matching lets people “curate” partners like playlists, inserting extra criteria (height, political leanings) that historically restricted assortative matching less severely.
- The trend is “a huge emancipation” for women — more freedom to choose — but also raises the question: how many opt out because the contemporary “rules of the game” are unattractive?
- “This is basically a massive experiment on the human population” — on large-scale social effects of AI companionship.
Actionable takeaways & recommendations
For policymakers
- Monitor demographic changes (fertility, singlehousehold rates) and adapt housing, pension, and social-care planning accordingly.
- Consider regulations for AI companionship (age verification, transparency on model changes, safety standards) to protect vulnerable users.
For technology companies
- Build safeguards around attachment and trauma risk (noticeable warnings for major model changes, data portability, opt-in/opt-out for memory features).
- Improve transparency: communicate limitations, update policies, and provide tools for parental control and verified age checks.
For individuals and communities
- Encourage real-world social skills and settings that allow formative dating practice (community activities, social spaces).
- Be cautious about relying solely on AI for emotional needs — seek human connections, therapy or community resources when possible.
For researchers
- Prioritize longitudinal studies to disentangle causation (e.g., does AI use increase loneliness or are lonely people more likely to use AI?) and to assess long-term societal impacts.
Bottom line
The rise in singlehood is complex and multi-causal: economic emancipation, pandemic disruption, and tech-enabled social changes all play roles. The consequences touch fertility, housing, politics and mental health. AI companionships are emerging as both potential mitigants of loneliness and new sources of risk — presenting urgent questions for tech companies, regulators and society about how to balance innovation with safety and human flourishing.