Chinese entrepreneurship attributes : a comparative GEM data analysis

Purpose: The purpose of this paper is to compare influential factors of entrepreneurial activities over time in China and to compare China with other selected countries. The data are collected from Global Entrepreneurship Monitor (GEM). The method used is decision trees and chi-square automatic interaction detector (CHAID) analysis, which isolates important factors and examines entrepreneurship predictor importance. Design/methodology/approach: The method used is decision trees and CHAID analysis which isolate important factors and examine entrepreneurship predictor importance. The original contribution of this paper is that this is the first time where artificial decision trees are applied to data to isolate factors that influence business startups and used across countries for comparative purposes. It is also the first application of this model to Chinese GEM. CHAID trees and predictor importance show the value of motivations of people who have already started businesses and shed light on how public policy can be influential in promoting entrepreneurship. Findings: Results indicate that solid knowledge and skills of how to start a business and knowing someone who has already started a business are the most important factors in China and in most of the selected countries. Fear of failure is becoming less important for Chinese entrepreneurs over the years from 2003 to 2012. Results show that countries, including China, have to enhance skill and knowledge education if they want to promote small business entrepreneurship as a policy. The findings support human capital theory. Research limitations/implications: The limitations of this study are due to using aggregated data from GEM surveys, which do not allow the authors to examine individual or household behavior. The authors do not know the variance and the distribution of responses to the questions asked and the locations in which the surveys were conducted. Another limitation is that GEM data do not report regional variations which can be modeled. For future work, the authors suggest more detailed data availability which will lead to isolating entrepreneurial problems and highlighting relevant attitudes important to entrepreneurs. Practical implications: Better data collection is needed at household and regional levels to understand business starts and to promote entrepreneurship. Social implications: Social implication of this research is to find out effective ways to increase entrepreneurial activities, therefore creating job opportunities and boosting economic growth. Educational programs will also decrease disparity of opportunity and incomes between different geographical regions in the country. The original contribution of this paper is that this is the first time artificial decision trees are applied to data to isolate factors that influence business startups across countries. Originality/value: The original contribution of this paper is that this is the first time where artificial decision trees are applied to data to isolate factors that influence business startups and used across countries for comparative purposes. It is also the first application of this model to Chinese GEM. CHAID trees and predictor importance show the value of motivations of people who have already started businesses and shed light on how public policy can be influential in promoting entrepreneurship. This research modeled the breakdown of reasons people would start a business by using GEM data surveys.