Analysing Japanese contractor activity on the basis of panel data

Purpose - The purpose of this paper is to examine the factors influencing defaults and exits from public works by prime contractors of small and medium-sized enterprises in the Japanese construction industry. By analysing the data for several years as panel data, this study determines the extent of influence of changes in company characteristics on the defaults and exits. Design/methodology/approach - Using construction company evaluation (Keiei Jikou Sinsa or Keisin) data and by employing the panel binary logit random effect model, this study empirically analyses the construction industry. Findings - This study shows that defaulting and exiting companies have different characteristics. The result shows financial performance indicators, non-financial performance indicators and Keisin scores to significantly affect defaults and exits. In particular, this study finds non-financial performance indicators, such as whether a firm draws insurance, to significantly affect its likelihood to default or exit and the influence varies on the basis of insurance type. Originality/value - The feature of this study is that its analysis focuses not only on defaulting companies but also on exiting companies, defined as those that stop operating as prime contractors for public works but otherwise stay in business. In contrast to existing research, this study distinguishes between defaults and exits and analyses the factors that influence a firm following one of these two outcomes. Moreover, although Keisin data are usually used to determine whether companies qualify to enter bids for public works, they can be applied for an attribution analysis of corporate defaults and exits.