Identifying foreign firms and South African multinational enterprises: CIT-IRP5 panel v4.0

Abstract

The identification of foreign firms and South African multinational enterprises (MNEs) in the CIT-IRP5 panel has proved to be a challenge for many researchers. The CIT-IRP5 panel contains variables indicating different thresholds that determine foreign ownership. The dataset also has variables that researchers can use to identify South African MNEs. Using the approaches employed by researchers who have attempted to identify foreign firms and South African MNEs in the data, four foreign firms and MNE indicators have been added to the CIT-IRP5 panel v4.0. This technical note documents the approach followed in the creation of each indicator. This note also highlights the possible company classifications in the data and fields on the ITR14 form that can be used to identify these classifications.

The guide to the CIT-IRP5 panel version 4.0

“…In this paper, we have presented the updated version of the CIT-IRP5 panel. We discussed substantial improvements incorporated based on the works of Budlender and Ebrahim (2020), Kerr (2020), and Kilumelume et al (2021), and showed that, despite changes in the underlying data, changes in vintages, and several form changes, the updated data largely conform to previous versions and external sources.…”

“…CIT-IRP5 version 4.0 includes a set of variables related to identifying firms with international connections. Kilumelume et al (2021) fully describe the classification of firms according to a strict foreign indicator, a broad foreign indicator, a domestic or foreign multinational indicator, and a foreign connection indicator.…”

Taxing top incomes in the emerging world: Economic impact under the microscope

“…As our sample period coincides with the implementation of various measures in South Africa and worldwide aimed at constraining income shifting by multinational companies to tax haven countries, we exclude these firms from our analysis to avoid picking up potentially confounding trends (this sample restriction is not decisive for any of the results presented below). Specifically, we drop all firms that are parent firms in South Africa (with ownership links to foreign countries) or subsidiary firms of foreign multinational companies located in South Africa, as identified following the broad multinational firm definition from Kilumelume et al (2021). The analysis also disregards large entities with more than 100 workers.…”