PensionReforms' summary and comments
The Japanese government is heavily involved in the running of pension schemes at Tiers 1, 2 and 3 but most particularly at Tier 1 (basic pension) and Tier 2 (compulsory, employment related).   There are already more than 35 million social security beneficiaries (about 30% of the total population) and that is increasing.  It seems that the government does not have a clear idea of who all these beneficiaries are.  The negative publicity around this issue contributed eventually to a change of government in 2007.

"Japan has several schemes of social security pensions among different sectors of the population. The oldest scheme dates back to 1884 and has a history of more than 120 years.  The newest one was established in 1961.  Before January 1997, pension identification numbers were issued to each participant independently from each pension program on the regional basis.  They used to be changed when the participant moved to other regions, to other companies or to other pension programs.  They were also changed when the participant acquired a new family name after his/her marriage or divorce.  This is mainly because there was no adding-up requirement of covered years among the different pension schemes.  Many Japanese, thus, were likely to have two or more pension identification numbers before retirement.  It is only in January 1997 that the unified pension identification number was introduced for all eligible persons in Japan."

When the unified system was introduced, the government discovered there were 300 million pension records but only 100 million eligible citizens.  So the government wrote to them all, asking them to report all their pension record numbers.  Initially, only 9.6% of them replied.  Over the years, that position has improved (pensions are paid only when records are reconciled) but, after a decade, there were still 50 million pension records that are described as "floating" because they are unrelated to a new "unified pension identification number".

There are apparently five main reasons for the administrative gaps:
-   human error, including fraud;
-   an assumption by citizens that bureaucrats knew what they were doing (in other words, a reluctance to check data);
-   on conversion from paper to punch-card digital records, converting names in Chinese characters was difficult because of technology limitations;
-   the absence of an integrated  tax and social security collection system allowed misreporting and fraud;
-   there was no overall monitoring organisation so no-one understood the scale of the problem.  Allied with that was the internal potential loss of face by the government administrators.

More than half of the 50.95 million "floating" records related to people over age 60 and since the problem became public, some progress has been made on resolving the gaps.  20% have since been "identified", 32% had already died or, apparently, weren't entitled to a pension and provisional matches have been made for another 15% (but only half of those have replied to a letter requesting confirmation).

Of the remainder, the authorities think they know who another 11% relate to and have still to investigate a further 23%.  That means "almost 17 million pension records remain unidentified."

"From April 2009, the SIA [Social Insurance Agency] began to send out social security pension statements ("orange letter") to all program participants annually. The statement includes pension information on the unified personal identification number, the insuree's name, the pronunciation of the name, gender, birth date and year, date of enrollment and/or leaving the program, the identification of the company where the participant worked/works, all records of monthly salaries and semi-annual bonuses he/she received, records of contributions and expected amounts of monthly pension benefits. The SIA strongly expects that upon receiving this statement, the participants will actively make responses, which will enable the Agency to integrate the pending pension records and to correct remaining errors in the SIA pension data base."

PensionReforms suggests the SIA is probably optimistic.  Perhaps the Japanese are more inclined to check this information more carefully than their Western counterparts (with whom PensionReforms is more familiar) but, faced with that welter of technical data, the recipients are more likely to file their "orange letters". And there are still 17 million unidentified records.  However, the suggested regular reporting to contributors will improve the quality of information held over time.

From 2010, a specialist Japan Pension Agency will be responsible for collecting and handling pension contributions. 

As the report concludes: "The trustworthy government with its competent and neat implementation is. the basis for any pension system." (File size 84 KB; 11 pp) 360