The Persistent Character of Japanese Employment: A Micro Data Analysis of Seikashugi and Its Impact on Japanese Workers

Seikashugi, the introduction of performance-related pay as an alternative or addition to ability-based pay (nōryokushugi), has been the most important change to the management of regular employees in Japan since the 1990s. As early initiatives at Fujitsu illustrated, the consequences of seikashugi can be rather disastrous (Jo 2004) and most firms have been relatively careful in its introduction (e.g. Keizer 2010). However, there has been little research on the actual impact on individuals' appraisal and wages (see JILPT 2005 for an exception). This paper aims to overcome this lacuna by using a unique set of micro-data on the appraisal and rewards system at a single firm and a survey among its employees concerning the system. It concerns a long-established construction firm with a global presence and about 4,000 employees. The data does not only include individual data on evaluation scores and remuneration but also aspects such as department, job rank, age/tenure, contract status, career track, educational background, type of recruitment, gender and home situation. Interviewees with three HR representatives and internal documents provide further information. The case study allows us to assess the impact of seikashugi on individual careers, the character of the Japanese firm, and the Japanese model of capitalism.

The data provide several important insights. First, important differences exist in the distribution of appraisal outcomes between departments and job grades with, for example, very high evaluation scores in the highest job grades. Previous research shows these findings to be representative (see Matsushige, Nakashima, Umezaki, Igawa & Kakizawa 2013). Two 'coping strategies' seem to shape outcomes: a 'centralisation effect', when evaluators tend to refrain from strong positive and negative judgements, and a 'leniency effect', when evaluators tend to assess performance positively. Both aspects can be related to a third, underlying factor: the invisibility of and thus the inability to assess performance. Variations across these three aspects can explain differences in performance outcomes and the paper discusses these differences in detail. However, a more fundamental issue concerns their implications for the viability of performance-related pay and they become clear when we include the survey data. For example, 'leniency' has a positive effect on people's trust in their performance evaluation and thus constitutes an important strategy to overcome distrust in the system. However, this creates a strong inflationary effect and undermines the long-term viability of performance-related pay.

The findings inform several conclusions. First, as systems of performance appraisal lose their bite, they will need to be 'renewed' every so many years. Secondly, the current support for performance-related pay, in and outside Japan, can be considered ideological as the data do not substantiate their superiority. Finally, the findings show light on the future of Japanese employment practices and the Japanese firm. The challenges of performance-related pay, especially in Japan where teamwork complicates appraisal and dissatisfied employees have few exit options, suggest that the impact of seikashugi remains limited and is unlikely to change the character of Japanese firms. It can also explain why some firms have reverted earlier seikashugi initiatives.