Work half a day, move sites, lose it all? The gaps in Korea's construction retirement system

What happens to a fractional workday — say, half a day — when a construction worker moves to a new job site before it can be counted?

The number of days actually worked is the foundation for calculating retirement deduction benefits for construction workers. When work records do not add up to a full day, the system carries the remainder over to the following month and combines it with subsequent records. The problem arises when a project ends or work is interrupted: those carried-over records may never be reflected in the final workday tally and simply disappear.

The Ministry of Employment and Labor said Sunday that an audit of the Korea Construction Workers Mutual Aid Association confirmed a systemic gap in which fractional workday records can be wiped out when employment is interrupted.

To understand how this problem arose, it helps to know how retirement deduction benefits accumulate for construction workers.

Day laborers and temporary workers at construction sites work differently from regular salaried employees who stay with one company for an extended period. When one project wraps up, they often move on to a different construction company or a different site.

The retirement deduction system was created specifically for these workers.

Under the system, construction business owners pay deduction contributions based on the number of workdays their employees report. The Korea Construction Workers Mutual Aid Association then accumulates those contributions for each individual worker. When a worker eventually meets the eligibility requirements and retires from the construction industry, the accumulated amount is paid out as a retirement deduction benefit.

The design allows a worker who moves from a site run by Company A to one run by Company B and then to Company C to have all those workdays added together when calculating retirement benefits.

The Korea Construction Workers Mutual Aid Association is a quasi-public institution that administers this retirement deduction program — along with welfare services and vocational training for construction workers — under the Act on the Improvement of Employment for Construction Workers and related laws.

The entire system rests on one key figure: the number of workdays.

At construction sites, labor is commonly measured in units called "gongsu," where one gongsu equals one full day of work. Depending on the nature of the work or the hours involved, a worker can end up with a fractional gongsu — a figure below 1.0.

How to handle these sub-one-day fractions was precisely what the audit flagged as a problem.

The mutual aid association's electronic card system handles fractional workday records by carrying them over to the following month and combining them with subsequent records. A batch-adjustment function processes certain ranges of fractional records together.

In practice, if a worker accumulates a sub-day fraction in one month but continues working into the next, the system can combine that fraction with later records and count it properly.

The problem is that there is no guarantee work will continue the following month.

Construction workers frequently move to a new site when a project ends or work dries up. If employment is cut off while a fractional workday is sitting in the carryover queue, that fraction may never make it into the final workday count.

According to the ministry's audit report on the Korea Construction Workers Mutual Aid Association, the association's electronic card system carried fractional workdays over to the following month and combined them there, while simultaneously processing certain ranges of fractional records through a separate batch-adjustment function.

The ministry said the criteria for carrying over, combining, and reflecting fractional workday records were not clearly defined.

Particularly problematic, the audit found, was the structure in which fractions already rolled over to the next month are erased outright when employment is interrupted.

The ministry concluded that two workers who log the same cumulative hours can end up with different reported workday totals and different retirement deduction balances — depending on whether they stayed at the same site continuously or had their employment interrupted at some point.

The scale of the problem was not trivial.

According to the audit report, the value of fractional carryover amounts that lapsed on completed projects came to 995 million won ($714,000) in 2024. In 2025, the figure was approximately 287 million won. The combined two-year total is 1.28 billion won.

Does that mean the 1.28 billion won represents retirement deduction money that construction workers never received?

Based on the audit report alone, that conclusion cannot be drawn with certainty.

The figure the ministry cited is specifically described as "the value of lapsed fractional carryover amounts based on completed projects." The report does not explicitly characterize it as retirement deduction money that went unaccumulated for workers.

The report also does not state how many workers were affected by the lapsed fractions, making it difficult to determine from the document alone exactly how much less was accumulated in retirement deduction benefits overall.

Even so, the systemic flaw is clear.

The ministry said the structure — in which previously carried-over fractions are erased when employment is interrupted — means that "even with the same cumulative working hours, the final reported workday count and retirement deduction accumulation can differ."

The retirement deduction system was designed from the outset to let construction workers, who rarely stay at a single site for long, combine workdays across multiple sites and receive retirement benefits accordingly. Yet the audit found that some work records already generated could disappear when a worker changes sites or has employment cut off.

The ministry agreed that reform is needed.

It called on the association to establish clear standards for handling fractional workday records and to improve the system so that fractional workdays already accrued are neither lost nor erased when employment is interrupted.

Ultimately, what this audit exposed goes beyond the question of who missed out on 1.28 billion won. The core issue is the structure itself: workers who put in the same hours can end up with different reported workday totals and different retirement deduction balances simply because their employment was interrupted — and that structure, the ministry said, needs to change.