Statistical modelling and analysis of business ownership in South Africa

Abstract

Business owners play a vital role in the social and economic development of South Africa by providing employment and services to the country’s citizens. South African infrastructure and community services are maintained by tax contributions collected from business owners. Furthermore, business owners also play the central part of circulating money in the county by providing services to the society in exchange of money. It is then evident that the success or failure of business owners directly affect the wealth of the country. In this study, the Quarterly Labour Force Survey (QLFS) and Survey of Employer and Survey Employers (SESE), both for 2017 from StatsSA was used, and the Generalised Linear Models (Multinomial Logistic Regression and Log-linear regression models) were applied to analyse and model business ownership in South Africa. The Chi-square statistic test from the descriptive statistics results showed that there is strong association between business ownership and the following categorical variables; gender, population group, marital status, age group, attended school, marital status, and province. The study utilised Multinomial Logistic Regression to identify factors affecting business ownership in South Africa. Gender, age group and attended school were three factors that are highly statistical significant with all their categorical levels having significant coefficients. Log-linear regression model was further used test if there was a significant interaction effect between four factors (own business, gender, population group and age group). The study found that only the 3-way effect interaction was significant, meaning that it had the high probability of improving the model than 4-way effect. Another objective of the study was to ii analyse the accessibility of finance by business owners. The study applied the Log-linear model using 2017 SESE data and found that the black population group dominates in terms of financial accessibility, the female gender also had a greater chance of getting access to loans than the male counterpart. The study recommends research on business ownership using the post COVID-19 data to investigate the effect of the Corona Virus pandemic on business ownership in South Africa using statistical methods.