The relationship between internal audit and foreign direct investment

ABSTRACT

This paper explores the influence of certain internal audit (IA) characteristics on the likelihood that FDI will be attracted. In making its investigation, the study uses data from the IIA CBOK database, and the World Bank database for 152 countries for the period 2012–2016. Ordinary least square methodology is used for analysis purposes, and the finding is that increases in FDI inflows result from effective IA. In particular, the most influential characteristic of IA in this respect is IA independence and competence, a finding which is valuable to academics and policy-makers who are keen to identify the drivers of FDI.

KEYWORDS:

• Internal audit
• foreign direct investment
• internal audit competence
• corporate governance
• national governance indicators

Acknowledgments

The data for the study was obtained from the Common Body of Knowledge in Internal Auditing (CBOK 2015). I thank the Internal Audit Research Foundation for its co-operation and for granting access to the CBOK 2015 Conditions of Anonymity and Confidentiality. However, whilst the data were provided by the Foundation, the views expressed in this study are those of the author and do not necessarily represent those of the Foundation.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Whereas Morrissey and Udomkerdmongko (2012) in their study construct a dummy variable for each NGI to show whether a country I characterized by high (1) or low (0) governance, this classification is underpinned by a country’s ability to score above the 50th percentile on the particular NGI concerned.

2. Other studies have used other variables as indicators of CG (board of directors’ independence, and AC independence and expertise). However, adequate data could not be obtained for these variables. Hence, due to the limited availability of data regarding CG, AC meetings’ frequency is used as one proxy of its effectiveness.

3. One valuable question to be posed in this connection is whether the correlations between IA characteristics are sufficiently high to cause problems of multicollinearity. However, the calculation of the Variance Inflation Factors (VIFs) reveals these to be in a comfortable position (below 10), thereby confirming no evidence of multicollinearity.

4. The IA variables are not highly correlated with the other variables used in the extended model, thus confirming the contention that these variables are important drivers of FDI inflows.

5. Regression was re-run using the average of the NGIs, and the results remained consistent showing significant impact on FDI.

6. When classifying countries to the two groups – developed and developing countries – the results suggested that the effects of IAINDEP and IACOMPET on FDI are significant at P < 0.01 in both groups, whereas the impact of INASIZE and IAMATUR are significant at P < 0.01 in developed countries and at P < 0.05 in developing countries, and FDI is positively correlated with IARESOURCES at P < 0.05 in both groups. These results provide additional support to previous findings that found differences of corporate governance between developed and developing countries.

7. Further bootstrapping tests were performed using the average of NGI (i.e., the five national governance indicators) and the combination of the five IA characteristics (IAMATURE, INASIZE, IARESOURCES, IAINDEP, and IACOMPET). The unreported baseline results of the combined components of IACHAR and NGI indicated that the findings remain consistent with small changes in the coefficients but no changes in the significance level of the other variables.

8. When controlling for year and country fixed effects, it is found that IA to have a beneficial effect on FDI and results remain consistent.

9. And 2SLS was re-run using total assets as an alternative instrument (as a proxy of firm’s size), which has been used in the study of the determinants of the use of IA (Carcello, Hermanson, and Raghunandan 2005). Moreover, IA outsourced (data analysis activities for internal audit is performed outside of IA department) was also used as another instrument, and when replacing total assets with IA outsourced, the results of 2SLS remain identical.

10. The average of the NGI indicators was also used when testing for the endogeneity. The results of the 2SLS remain stable and the significance levels of the other variables remain the same.

11. When testing whether CG and NGI moderate the effects of IA on FDI, again education was added as an additional indicator of IACOMPET and annual training hours was replaced with average training, and the results of the interaction effects were identical to the main results.

12. When performing the moderation tests between components of IA and NGI indicators, the interaction between QUALITYPS and IA variables shows a significant impact on FDI although the results of Model 2 (Table 5) show no significant association between FDI and QUALITYPS.

13. When using IACHAR (components of IA characteristics), the results showed that the effect of IACHAR on FDI is significant in all income groups.

14. After including these additional variables, the previous tests (the additional tests, moderation, and mediation tests) were performed again. Whilst there were slight changes in the coefficient, the level of significance remained the same as the variables.