On a humid Tuesday morning in August 2024, one hundred women walked out of Incheon International Airport into a country that had spent two years arguing about whether they should come. They were Filipina caregivers. Each had completed 780 hours of state-approved training. Behind them stood a government promise that the Korea foreign caregiver program would finally make child care affordable in the world’s lowest-fertility nation.
Two years later, 82 of them remain. The rest went home. The program that was supposed to grow to 1,200 workers was quietly killed in December 2025, and no additional visas will be issued. Meanwhile, the same policy machinery is now being pointed at a problem roughly a thousand times larger: Korea will be short 116,000 elderly care workers by 2028.
So what actually went wrong? The short answer is not the one most people expect. The pilot did not fail because Koreans rejected foreign workers. Instead, it failed because the arithmetic never worked, and because a single word in a bilateral agreement was translated two different ways.
August 6, 2024: The Korea Foreign Caregiver Program Lands
The idea belonged to Seoul Mayor Oh Se-hoon, who first floated it in 2022. His pitch was simple. Hong Kong and Singapore had built entire middle classes on the backs of affordable live-in help. Korea, facing a birth rate that had collapsed below 0.8, could do something similar and let exhausted parents stay in the workforce.
Getting from pitch to tarmac took two years of negotiation with Manila. Eventually, the Ministry of Employment and Labor and the Seoul Metropolitan Government settled on a pilot: 100 Filipina workers, admitted on E-9 visas, deployed to households across the capital. Each worker arrived with certification and language preparation. In addition, Korean labor law would apply in full, including the minimum wage.
That last detail sounded like a protection. In practice, it was the pilot’s death sentence.
Trouble surfaced almost immediately. Within two weeks, two participants failed to return to their assigned dormitory. Both eventually flew back to the Philippines in October. Korean media treated the disappearance as a scandal about the workers. Labor researchers, however, read it as a signal about the conditions waiting for them.
The timing mattered too. Korea’s births had just begun ticking upward after nine straight years of decline, and the government wanted a visible answer for young parents. Consequently, the Korea foreign caregiver program launched under political pressure rather than after a costing exercise. Anyone tracking Korea’s maternity care system will recognize the pattern: demand recovers faster than the infrastructure meant to serve it.
The Price That Was Never Cheap
Here is the number that broke the Korea caregiver pilot. A household using the service for 40 hours a week paid roughly 2.43 million won a month in 2024. From March 2025 onward, that figure climbed to about 2.92 million won, or close to $2,090. The hourly rate rose from 16,800 won to 18,900 won.
For context, the average monthly cost of hiring a domestic caregiver in Korea in 2023 was about 2.64 million won. In other words, the government imported workers from 4,000 kilometers away and produced a service that cost more than the domestic option.
Several forces pushed the price up. First, Korea’s minimum wage applied without exception. Second, workers employed longer than a year became eligible for severance pay, which added roughly 497,000 won a month to the bill. Third, the program ran through private platforms rather than direct hiring, so intermediary margins stacked on top.
As a result, the central premise evaporated. Korea Herald reporting put it bluntly: the pilot was sold as cheap child care, and it was not cheap. Ordinary dual-income couples, the exact group the policy targeted, looked at the invoice and walked away.
Consider a typical Seoul household earning 6 million won a month after tax. Rent or mortgage takes a third. Then comes 2.92 million won for care. Nothing remains. For that family, the Korea foreign caregiver program was never a real option, no matter how sympathetic the press coverage was.
The March 2025 handover deserves its own note. Until then, the city had run operations directly. Afterward, private platforms took over placement and housing, and costs rose again. Naturally, each additional layer needed a margin. Meanwhile, the workers absorbed the housing charge while households absorbed the fee.
Half the Average Wage
Meanwhile, the workers themselves were not getting rich. A study published in the Korean Journal of Immigration Policy and Administration surveyed 21 of the participants, most in their twenties and thirties. Their average gross pay over the first six months came to 1.92 million won a month. Korea’s average monthly wage that year was 3.74 million won. The caregivers, therefore, earned 51 percent of the national average.
Take-home pay told an even sharper story. After deductions, the average was 1.18 million won. Some workers cleared less than a million. Housing alone consumed up to 520,000 won a month, because dormitory costs were charged back to them.
Run the math on hours worked and the effective rate lands at 9,860 won an hour. By comparison, that sits 27 to 35 percent below what Korean domestic workers and child care helpers typically earn. So households paid a premium while workers received a discount. The gap disappeared into housing charges, platform fees, and administrative overhead.
The satisfaction scores followed the money. Households rated the service 4.07 out of 5. Caregivers rated their experience 3.47. Notably, that spread is the whole policy in one statistic.
Remittances made the squeeze concrete. A Filipina worker sending money home from Hong Kong keeps most of her salary, because her housing and food come from the employer. Under the Korean model, by contrast, she paid Seoul rents out of a Seoul-adjusted wage. Follow-up research found participants performing unpaid tasks on top of that, which pushed the real hourly figure down further. Word travels quickly through migrant networks. Within a year, the Korea foreign caregiver program had lost whatever recruiting advantage it started with.
Caregiver or Housekeeper? One Word, Two Meanings
The memorandum of understanding between Seoul and Manila used a specific term: caregiver. Under that document, the women were to help children and pregnant family members with daily activities. Household chores were to be “incidental and light.”
Korea’s own paperwork said something else. Both the national government and the city classified the arrivals as domestic workers, a service-sector category. Attorney Choi Jeong-gyu of Wongok Law Firm flagged the mismatch directly, noting that the two governments were, in effect, describing different jobs.
Predictably, the ambiguity got filled by whoever was standing in the kitchen. Survey data showed participants spending 64.8 percent of their time on child care and 30.2 percent on housework. Beyond that, workers reported unpaid extras: general cleaning, dishwashing, pet care, and English tutoring for the children. At a Seoul Metropolitan Council debate in June 2025, testimony described schedules running from 8 a.m. until midnight.
None of this was hidden. Rather, it was structural. When a contract defines duties loosely and the worker cannot change employers, scope creep is not a risk. It is the default.
Who Actually Used the Foreign Caregiver Program
The demographic data deserves more attention than it received. According to a labor ministry survey, 73 percent of subscribing households had combined incomes above 9 million won a month. Furthermore, 40 percent lived in Gangnam, Seocho, or Songpa, the three wealthiest districts in Seoul.
Consider what that means. A policy justified as relief for struggling young families delivered subsidized convenience to the top income decile. The families who most needed help could not afford 2.92 million won a month. The families who could afford it did not need a national demographic emergency to justify their hiring decision.
This pattern is familiar to anyone who follows Korean family policy. Similar dynamics show up in the English kindergarten market, where officially capped tuition still produces waiting lists in the same three districts. Likewise, the broader kids economy has bent toward premium spending on fewer children rather than toward volume.
The E-9 Trap
Even so, the deepest problem was the visa. Korea’s E-9 permit, issued under the Employment Permit System, is designed for non-professional labor. It restricts workers from freely choosing a workplace. Changing employers requires permission, and residency is tied to continued employment.
Consequently, a caregiver asked to scrub floors at 11 p.m. has three options: comply, negotiate from a position of zero leverage, or leave the country. Professor Lee at Jeju National University framed the trade-off precisely. Koreans avoid care jobs because the conditions are poor. For migrants, those same conditions come bundled with residency insecurity.
Administrative structure made things worse. Responsibility was split three ways among the central government, the Seoul city government, and private platforms. Consequently, no single body had clear authority to intervene when disputes arose. One platform executive explained that because visa operations sat with the Employment Ministry, the service could not be continued independently.
Readers who want the wider picture on how these permits work should see our breakdown of the 2026 work visa overhaul, which consolidates 39 categories into three.
How Hong Kong, Singapore and Taiwan Do It
Korea did not invent this policy. Its neighbors have been running versions of it for decades, and the contrast explains a great deal.
Hong Kong hosts roughly 360,000 foreign domestic helpers. The city sets a Minimum Allowable Wage rather than applying its general labor floor. Since September 30, 2025, that rate has been HK$5,100 a month plus a HK$1,236 food allowance. Total monthly employer cost therefore lands near $810.
Singapore employs about 316,900 migrant domestic workers as of December 2025. Employers there pay a monthly levy of S$300, reduced to S$60 for households with a young child, an elderly member, or a person with a disability. Meanwhile, salaries are set by sending-country floors rather than Singaporean law. Filipina workers start around S$650, tracking Manila’s US$500 minimum. Total employer cost typically runs $550 to $780.
Taiwan counts roughly 233,000 migrant social welfare workers, about 31 percent of its entire migrant workforce. Average monthly compensation was NT$20,209 in mid-2022, or about $640.
Now place Korea beside them. The country hosted 6,644 foreign care workers in 2024. That is not a rounding error against Taiwan’s 233,000; it is a different order of magnitude entirely. Moreover, Korea’s per-worker cost to households runs two to four times higher than any of the three.
The reason is not generosity. It is architecture. Hong Kong, Singapore, and Taiwan all built separate legal categories for domestic care, with their own wage floors, their own levies, and their own contract rules. Korea, by contrast, tried to run migrant care through a standard labor framework while hoping the price would somehow come out low. Two things cannot both be true.
Each neighbor pays a price for its choice, and the trade-offs are worth naming. Hong Kong requires live-in arrangements, which has drawn sustained criticism from labor groups over privacy and working hours. Singapore’s levy system gives the state a revenue lever but leaves wage-setting to embassies. Taiwan’s caregivers work an average of ten-hour days, often without defined limits. In short, none of these systems is a model of worker protection. What they do have is a coherent design, and that is precisely what the Korea foreign caregiver program lacked.
That design gap sits at the center of an unresolved argument. In 2024, the Bank of Korea published a report suggesting foreign care workers be exempted from the minimum wage. Migrant unions and women’s organizations condemned it immediately. The debate never resolved, and Korea has still not ratified ILO Convention 189 on domestic workers. Instead of choosing, policymakers ran a pilot that avoided the question and let the numbers answer it.
December 23, 2025: A Quiet Ending
There was no press conference. On December 23, 2025, the government announced that no further visas would be issued for the scheme. The planned expansion to 1,200 workers simply stopped existing. Around the same period, Korea cut its overall E-9 quota from 130,000 to 80,000.
By then, attrition had done most of the work. Of the original 100, seventeen had returned to the Philippines and one had moved into hotel cleaning. Approximately 82 remained, serving 133 households in Seoul. Korea JoongAng Daily’s post-mortem collected the verdicts from academics and lawyers, and the consensus was uncomfortable. Jung Jae-hoon of Seoul Women’s University summarized it as Korean society’s structural limits showing up inside a small program.
Still, calling the pilot a failure requires care. It failed as a cost-reduction policy. As an experiment, however, it worked perfectly. It produced clean evidence about what Korea’s labor architecture does to imported care work, and that evidence arrived just in time for a much bigger decision.
The Next 116,000
On August 16, 2026, the Korea Herald ran a story with a question in the headline: who will care for Korea’s elderly? The numbers underneath it are the reason the caregiver debate is not over.
Korea will need about 800,000 care workers by 2028 and expects to have roughly 690,000. That leaves a gap of 110,000 to 116,000 workers. Looking further out, the country needs 990,000 additional caregivers by 2043.
The existing workforce is itself aging. Of the 657,104 care workers counted in 2024, about 66 percent were 60 or older. Additionally, roughly 2.3 million people hold caregiver certifications and do not work in the field, which tells you what the job pays and how it feels.
Government strategy is already shifting. In March 2025, the Justice and Health ministries approved 24 universities to run degree programs for international students seeking caregiver qualifications, with visa incentives designed to encourage long-term settlement. A two-year pilot began in 2026. Policy researchers have recommended opening E-9 entry into care training and creating an E-7 skilled pathway afterward.
Yet the same warnings keep surfacing. Kim Ok-nyu, a professor studying the sector, pointed out that care work is already a job Koreans avoid, and that migrants will not automatically want it either. Lee Kyu-yong of the Korea Labor Institute was blunter, arguing that the sequence is wrong: reform the working conditions first, recruit second.
For investors, this is where the story turns commercial. A structural shortfall of a million workers reshapes nursing home operators, staffing platforms, and the automation market. We have covered the demand side in Korea’s silver economy and the technology response in care robots. Neither substitutes for the labor question. Both get more valuable the longer it goes unanswered.
There is a domestic supply option too, and Korea is already testing it. Raising the retirement age would keep experienced workers in the system for longer, and 2.3 million dormant certificate holders represent a large idle pool. Even so, the arithmetic does not close. Care demand grows faster than any plausible reactivation rate, which is why migration keeps returning to the table despite the Korea foreign caregiver program and its unhappy ending.
What Foreign Residents in Korea Should Know
If you live in Korea and were hoping the foreign caregivers in Korea scheme would return, plan without it. Nothing in current policy suggests a household-level program is coming back soon. Practically speaking, the realistic options today are the government’s subsidized Idolbom sitter service, private sitter platforms, standard daycare, and the private kindergarten market. Costs vary enormously across those four.
Idolbom is the cheapest route, because subsidies scale to household income, though waiting lists in central Seoul are long. Private platforms cost the most and move the fastest. Public daycare sits in between and rewards early registration more than anything else. Above all, avoid the informal market. Hiring an undocumented worker exposes both sides to real legal risk, and the worker carries almost all of it.
If you work in migration, labor policy, or elder care, watch the university pipeline instead. The 24-campus program is where Korea’s next attempt is being built, and it is being designed to avoid exactly the failures documented above: better-defined roles, a credential pathway, and residency that does not evaporate when a placement ends.
And if you are simply trying to understand the country, hold onto this. Korea did not reject the Korea foreign caregiver program out of xenophobia. It applied its own labor standards honestly, discovered that honest application makes care expensive, and shut the program down rather than resolve the contradiction. The elderly care shortage will not allow that kind of postponement twice.
One hundred women landed at Incheon in 2024. Eighty-two are still here. The next cohort will be measured in the hundreds of thousands, and Korea has about two years to decide what kind of jobs it is offering them.
