The relationship between company size and CEO remuneration: A scaling perspective

Abstract

The first objective of the study is to empirically test a number of company size determinants’ significance as size proxies in benchmarking CEO remuneration for different sectors of Johannesburg Stock Exchange (JSE)-listed companies. The second objective is to investigate an issue that has not been examined in previous studies, namely the extent to which companies are able to linearly scale their CEO remuneration and company size without changing the remuneration-to-size ratio. To fulfil the first objective, data extracted from the McGregor BFA database were obtained for 2013, where 244 companies in four sectors, i.e. financial, manufacturing, minerals and services, are analysed using descriptive statistics and simple regression analysis. From the results obtained, to fulfil the second objective, a data envelopment analysis (DEA) model is built to estimate the technical and scale efficiencies of 231 companies. A hypothesis test was helpful to find that the following determinants can be used as proxies for company size: total assets (including intangible assets); market value of assets; total equity; market capitalisation; revenue; and total cost. The confidence level to which the null-hypothesis is rejected leads to the conclusion that those determinants are on their own suitable proxies that make further investigations into joint determinants unnecessary. Furthermore, the study concluded that the majority of companies are not able to linearly scale their CEO remuneration and company size without changing the remuneration-to-size ratio. Therefore, the conceptual theory of scaling is to a great extent rejected, since only nine of 231 companies in the sample investigated could achieve economies of scale. The paper is organised as follows: Section I provides the gap of missing knowledge in the literature as well as the conceptual framework of the study. The data and methodology are described in Section II, after which the results and a discussion thereof are provided in Section III. The study is finally concluded in Section IV.