Corruption and Corporate Cash Holdings

ABSTRACT

In this study, we examine the relation between corruption and corporate cash holding policy using data from 35,288 firms in 44 countries. The results indicate a non-linear relation between corruption and corporate cash holdings with a negative coefficient of the cubic term of corruption. Specifically, in developed (frontier) countries with low (high) level of corruption, we find a negative relation, whereas in emerging countries with moderate level of economic development and corruption, we find a positive relation. Thus, we suggest that corruption is an important variable that affects corporate cash holding policy by the level of economic development.

KEYWORDS:

• Cash holdings
• financial policy
• corruption
• emerging markets

JEL:

• G30
• G32
• G38
• F50

Acknowledgments

The authors would like to acknowledge the financial support provided by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2020S1A5A2A01044408).

Notes

1. This previous study similarly covers 47 countries; however, the number of firm-year observations is limited to 68,376 from 1996 to 2007, because of data availability.

2. In our sample, the corruption index of the United States is 2.57 on average, a very low level of the total score (10). Smith (2016) shows that the threat of expropriation by corrupt officials exists even in the U.S., which is not typically viewed as a corrupt country.

3. If the original value of the corruption index is three (total score is ten), we change it to seven (total score is also ten).

4. For example, in 2019, the ratio of Argentina is 0.927, which means the number of respondents who answered six points or more represents 92.7% of the total respondents.

5. We try to include all variables of cultural characteristics, that is, uncertainty avoidance, individualism, power distance, masculinity-femininity, long-term orientation, and indulgence versus restraint. However, the multicollinearity problem among cultural variables in our sample is a serious one; thus, the VIFs are higher than ten. Therefore, we only add uncertainty avoidance (UAI), which is the most important cultural variable in our regression model (Ramirez and Tadesse 2009). Moreover, we drop UK legal origin, as it has a high correlation with UAI (correlation coefficient = −0.909).

6. The coefficient of TI is negative and statistically significant at the 1% level; therefore, there exists a negative relation in the sample with very low levels of corruption in developed countries. The coefficient of TI$\times$Middle is positive and statistically significant at the 1% level. This indicates that cash holdings in the middle samples increase in this range. The coefficient of TI$\times$Extreme is negative and statistically significant at the 5% level, indicating that cash holdings in extreme samples are lower than those in other samples.

7. In the unreported sub-sample results, in the developed countries sample, a U-shaped relation exists between corruption and cash holdings; however, in the emerging and frontier countries sample, we find an inverted U-shaped relation. These results also support our primary results on Table 3.

8. The weak instrument variable test is Kleibergen and Paap (2006)’s Wald test of the relation between state corruption and the instrument variables.

9. Owing to high correlations with the colonial dummy, we exclude inflation and UK legal origin in our regression models in Model (5) and Model (6), in Table 5.

Additional information

Funding

This work was supported by the National Research Foundation of Korea [NRF-2020S1A5A2A01044408].