BLOG

MATCHMAKING WITH AI

16.10.21

In December 2020, the BBC published an article detailing how Japan’s government plans to fund artificial intelligence matchmaking schemes to combat the nation’s tumbling birth rate. 


The state taking an active interest in the love lives of inhabitants feels like something from a Black Mirror episode, but with birth rates falling to a record low it also seems inevitable.

According to the article:


“Japan's population is projected to fall from a peak of 128 million in 2017 to less than 53 million by the end of the century.”

It continues:


“[In 2021] the government plans to allocate local authorities 2bn yen ($19m, £14m) to boost the birth rate, reported AFP news agency.


“Many already offer human-run matchmaking services and some have introduced AI systems in the hope they will perform a more sophisticated analysis of the standardised forms where people submit their details.”


Rather than focus on typical criteria - such as age, looks and income - AI-based systems will favour a so-called “emotional quotient”: a term referring to similar personalities, values and emotional intelligence. It makes sense. Not all 30-something, university-educated folk with a steady income are compatible. Nor are unemployed, 21-year-olds living with their parents. It takes a lot more than circumstances to form a meaningful connection.


But is AI matchmaking really the best way forward?


Sachiko Horiguchi, a socio-cultural and medical anthropologist at Japan’s Temple University, doesn’t think so. According to the BBC:


“[Horiguchi] thinks there are better ways for the government to bump up the birth rate than subsidising AI matchmaking - such as helping young people earning low wages.


“She pointed to a recent report which suggests a link between lower income levels and the loss of interest in romantic relationships among young Japanese adults.


“‘If they're not interested in dating, the matchmaking would likely be ineffective,’ Dr Horiguchi told the BBC. ‘If we are to rely on technologies, affordable AI robots taking over household or childcare tasks may be more effective.’”


The article concludes:


“Analysts have long pointed towards the lack of support for working mothers in Japan, where there are strong expectations women will do all the housework and raise children alongside doing their jobs.”


Japan may be the first country to venture into government-funded AI matchmaking, but there are others working to combat their decline in population. In January 2020, the BBC reported:


“Countries need to have a birth rate of at least 2.1 children per woman to sustain the population, but the average figure in Europe is about 1.59.


“According to the UN, two-thirds of countries in Europe have introduced measures to increase fertility rates, from baby bonuses and tax incentives to paid parental leave, with varying degrees of success.”


President Putin hopes to raise Russia’s birth rate to 1.7 children per woman (from 1.48 at the time of the report). Under the proposals, maternity benefits would be extended to first-time mothers for the first time, welfare benefits would be paid to low-income families with children aged three to seven, and free school meals would be provided for the first four years of school.


The Italian government has tested similar financial incentives to encourage couples to have more children, including an €800 payment per couple per birth (launched in 2015). Despite this, Italy still has one of the lowest fertility rates in the EU. The article explains:


“In the case of Italy, the failure of the incentive may be related to the fact that one-off payments fail to address underlying issues such as social attitudes - an important factor in a country with less than 50% of women in work - or large levels of emigration.”


Prior to the COVID-19 pandemic, France boasted one of the highest fertility rates in the EU with 1.92 births per woman (according to 2017 World Bank figures):


“The country has extensive social policies, which provide subsidised child care for younger children and a generous benefits system ‘especially for larger families’, according to the European Commission.


“Families with two or more children receive benefits of at least €131.55 per month and means-tested grants are available, including a payment of €944.51 given at the birth of each child for eligible families.”


Sweden has seen similar success.


“Generous family and social benefits may play a role in their success. Swedish parents are also entitled to 480 days of paid parental leave to share between them, with men claiming about 30% of all leave.


“Childcare is also subsidised and working hours are lower in Sweden than in many other countries. In 2018, the average Swede worked 1,474 hours, about 500 hours fewer than the average Russian.”


But even Sweden isn’t exempt from decline – especially when you throw a deadly pandemic into the mix.

In a fascinating article for the New Statesman, Sophie McBain explores how the “horror and uncertainty” of the COVID-19 pandemic had a dramatic contraceptive effect on the population:


“Birth rates tend to fall in the immediate aftermath of crises – flu pandemics, recessions, natural disasters – but many features of the coronavirus pandemic are unique. Extended lockdowns have made it hard for single people to find partners, or for long-distance couples to meet. The strain on working parents who have been home-schooling or looking after small children has been immense, making it more likely that these families will abandon or postpone plans to have another child. 


"The harrowing experiences of pregnant women who have had to labour or miscarry alone, and the isolation experienced by new parents may have caused some onlookers to delay their plans to start a family – certainly, some have told me as much. Some will find that, by the time they feel ready, they are no longer able to conceive. Fertility treatments such as IVF have been delayed. The stress and unhappiness of pandemic parenting can have diffuse effects.


"I spoke to a woman in her mid-twenties who said that witnessing these struggles second-hand had convinced her that she never wanted children: she didn’t want to take the risk that there would be another pandemic and that she’d end up as miserable as her friends with kids.”


And the figures speak for themselves:


“In the US, the fertility rate fell by 4 per cent in 2020, to the lowest on record. Italy’s birth rate has dropped to its lowest level since unification in 1861; together with a high Covid-19 death toll, this has caused a drop in population equivalent to a city the size of Florence. In France birth numbers have dropped to their lowest since the Second World War; in Japan and South Korea there have been record lows. The number of births in China dropped 15 per cent in 2020; after decades of maintaining a one-child policy, replaced with an allowance for two in 2016, the government announced in May that women could now have three children.


“These figures are striking taken in isolation, but represent an acceleration in a decades-long trend – one that will completely reconfigure the global economy, the international balance of power, and our intimate and personal lives. It will require fundamental social change to accommodate the diminishing size of the tax-paying, economically productive population, as well as the rising number of older people requiring pensions and social care.”


When put like that, it’s clear to see why Japan’s government are turning to technology to tackle the decline. Artificial intelligence, in this respect anyway, is an unknown quantity. There isn’t hundreds of years’ worth of data to rely on. There is data, however, detailing how past incentive schemes have tried and failed, so why not give it a go? And if unlucky-in-love singletons find companionship and happiness as a by-product of a government-funded baby-making drive, then what’s the harm?


The cynic in me worries about government interference, privacy risks, catfishing, data leaks and the like. The romantic in me hopes for the happily ever after. Only time will tell…