20% of data in labor ministry survey on working hours dropped due to flaws

TOKYO -- The Health, Labor and Welfare Ministry has found flawed figures on the working hours of regular employees at 966 workplaces covered by a 2013 survey, ministry officials said.

The discovery prompted the ministry to delete data at these workplaces from the survey results. This cleanup followed a similar ministry move over flawed data on work hours of those employed under the discretionary labor system. The total number of workplaces whose labor hour data have been dropped due to irregularities has now come to 2,492 -- about 20 percent of workplaces covered by the survey.

Health, Labor and Welfare Minister Katsunobu Kato apologized for the irregularities. "We must humbly reflect on the problem," he told a House of Representatives Health, Labor and Welfare Committee session on May 15. Nevertheless, ministry officials maintain that the survey outcome is statistically reliable, noting that the survey still covers over 9,000 samples.

Opposition parties disagree. They argue that the basis for deliberations on a government-sponsored bill to reform the way people work -- the so-called work-style reform bill -- has been shaken, pointing out that the bill is based partly on the survey outcome.

So what were the irregularities?

In the survey conducted between April and June 2013, labor standards inspectors visited 11,575 workplaces across the country and surveyed employees' overtime hours and other data. The inspectors checked both the work hours of regular employees, and those of people working under the discretionary labor system. Discretionary workers are rewarded based on fixed overtime work instead of actual labor hours.

Based on the initial results, the Health, Labor and Welfare Ministry determined that discretionary employees work for an average of nine hours and 16 minutes a day. The figure for regular workers came to nine hours and 37 minutes per day.

These figures were the basis of the argument by Prime Minister Shinzo Abe in his Diet testimony in January that the data showed discretionary workers work shorter hours than regular workers. The premier was pushing the work-style reform bill, which included the expansion of the discretionary labor system.

But opposition legislators, who think the bill would remove limits on work hours without proper compensation, were not convinced. They pointed out that the survey outcome contained abnormal data -- with daily work hours exceeding 24 hours, or the monthly work hours of some employees being shorter than their weekly work hours. Problems involving data comparing discretionary labor hours and regular hours were also discovered.

After tallying the figures again following the deletion of the flawed data, the health ministry determined that the working hours surveyed were shorter than the initial evaluation. Regular workers' average daily overtime hours were revised downward from one hour and 37 minutes to one hour and 33 minutes. Their monthly average overtime hours were lowered from eight hours and five minutes to six hours and 55 minutes.

After these revisions, the government ended up deleting clauses related to the expansion of the discretionary labor system from the work-style reform bill.

In the Health, Labor and Welfare Ministry's attempts to explain this failure, officials said it resulted partly from problems with the questionnaire design and guidelines for filling out survey sheets. A ministry official added that the survey was the first of its kind for some labor standards inspectors, and it was difficult for them to understand how the survey should be carried out. The ministry also acknowledged that it failed to thoroughly check the data when initially tallying it.

"We didn't have enough time and ended up conducting a sloppy survey," recalled an inspector at a labor standards inspection office in eastern Japan who surveyed about 10 workplaces.

(Japanese original by Shunsuke Kamiashi, City News Department)