For most of the post-war period Japan ran an unusual experiment: a wealthy, ageing, labour-short economy that declined to solve the shortage the way every comparable country did. Automation absorbed some of it. Longer working lives absorbed more. Women entering the workforce in larger numbers absorbed a great deal. What remained was met by a set of programmes that brought in foreign workers while carefully not describing them as immigration.
The most consequential of those was a technical training scheme, established on the premise that workers would come, acquire skills, and return home to use them. In practice it supplied labour to agriculture, food processing, construction and shipbuilding, and the training content was frequently notional. Everyone involved understood what it was for. The fiction was maintained because the alternative required a political conversation nobody wanted to have.
The practice moved first, and the policy followed at a distance
That conversation is now happening in slow motion, largely because the arithmetic stopped being avoidable. The working-age population has been shrinking for three decades, the shrinkage has accelerated, and the sectors most exposed are the ones with the least capacity to automate. A convenience store can install a self-checkout. A care home cannot install a care worker.
The reforms of recent years have created routes that look much more like conventional immigration: longer stays, family accompaniment, pathways to permanent residence for workers in designated sectors. The number of foreign workers has roughly tripled from where it stood at the start of the century, and in some prefectures the share of the manufacturing workforce is now high enough that a plant closure and a visa decision are the same conversation.
What has not kept pace is the infrastructure that makes migration work for the receiving society. Language provision is thin and unevenly funded. Housing in the industrial towns where demand is highest was built for a population that is no longer there, an inversion of the company town returning with better intentions. Local governments carry most of the integration cost and almost none of the fiscal benefit, which is a familiar misalignment and a reliable predictor of local resistance — the same split that makes disaster costs a fight between states and the centre.
The competitive picture is the part that should concern Japanese employers most. For twenty years the country could assume that workers from lower-income neighbours would keep arriving because the wage gap was decisive. That gap has narrowed considerably, several origin countries are ageing themselves, and the destinations competing for the same workers — Korea, Taiwan, the Gulf, increasingly parts of Europe — have been more direct about offering settlement rather than rotation. A programme built on the assumption that workers will leave is a poor offer in a market where the alternative is to stay.
Employers have noticed, and the ones adapting fastest are doing what employers everywhere do when a labour supply tightens: they are shortening the distance between hiring and usefulness. Firms that once relied on the training fiction are now running genuine onboarding, paying for language instruction, and promoting from within a workforce they previously treated as temporary. It is the same discovery driving the federal programmes that finally show wage gains and the same logic behind hiring for skills rather than credentials — that the constraint is rarely the applicant and usually the path from applicant to productive.
The unresolved question is whether the policy will ever say plainly what the practice already does. Japan has arrived, by increments and without announcing it, at being a country of immigration. The institutions have not been told.
Topics worldlabordemographics
