This module is a resource for lecturers  


Gender differences in corruption


Before exploring the emerging literature on gender differences in corruption, a brief note is included on the need to think critically about some of the common assumptions in this field. Having this "caveat" in mind may contribute to a deeper understanding of the relevant literature.

Avoid essentialist, reductive thinking

In exploring questions concerning the influence of gender on corrupt acts, it is important to avoid essentialist, reductive thinking. This means avoiding understanding a particular gender as defined by a set of necessary and sufficient features that all persons of that gender share. Any particular gender is better thought of as a " family resemblance concept", where all members share a number of features in common, but do not necessarily all share any one particular feature in common (Wittgenstein, 2001). This indicates that apparent causal relationships between gender and certain outcomes will not hold for all members of that gender, given the vast differences between individuals within any gender. Individual personality matters, and so do contextual factors like class, race, vulnerability and poverty levels. In fact, the interaction of those contextual factors can lead to effects that go beyond the total sum of their individual effects. The analytical framework of intersectionality is used to describe how other factors play into gender discrimination, particularly race, compounding its effects (Crenshaw, 1991). For a further discussion and a class exercise on intersectionality see Module 5 and Module 9 of the E4J University Module Series on Integrity and Ethics, and Module 9 of the E4J University Module Series on Crime Prevention and Criminal Justice. See also this entry in the Stanford Encyclopedia of Philosophy by Allen (2016).

Avoiding essentialist, reductive thinking also includes being aware that gender (or sex) will rarely be the sole or even primary determinant of a person's behaviour (Debski and others, 2018). Arguably, in most situations it simply correlates with other underlying explanations of differences between genders. Students should be encouraged to critically analyse and debate gender and corruption without resorting to essentialist, reductive thinking. For a related discussion on the issues associated with essentialism, see Dzubinski and Diehl (2018). Furthermore, caution should be exercised when considering evidence seeming to support claims such as that "women are less corrupt than men", "women are more adversely affected by corruption" or "women in public office will - by virtue of their gender - advance and prioritize women's rights or the rights of vulnerable groups". The role that gender can play in preventing corruption in the long term requires further research in numerous contexts.

Another matter to bear in mind when discussing gender and corruption is that much of the work on gender relies heavily on sex-related data. Breaking down data by gender (gender-disaggregation) is challenging because gender is difficult to measure and quantify, and because the value of distinguishing it from sex is sometimes neglected. Sex-disaggregated data are therefore often treated as a rough proxy for gender-disaggregated data (although some countries are starting to collect data on both gender and sex - see examples discussed here). However, as demonstrated below, sex-disaggregated data can conceal nuanced gendered processes underlying corruption.

The state of research on the link between gender and corruption

Research on gender, sex and corruption emerged in the early 2000s with two studies from the World Bank, which reported a correlation between low levels of corruption and more women in government. The first study, conducted by Dollar, Fisman and Gatti (1999), explored a large cross-section of countries and found that a higher proportion of women in parliaments were associated with lower levels of corruption. This relationship remained unchanged when various measures of corruption were used and after the influence of several variables related to both gender and corruption, such as civil liberties, income and education, were removed from the statistical analysis.

The second study, Swamy and others (2001), found a similar correlation. On examining different kinds of female participation more closely, the authors found that countries with more women in parliament, in ministerial posts and in the labour force, were less likely to experience corruption. The authors found further support for their findings through an analysis of micro-survey data of business firms in Georgia. In particular, they found that those companies owned or controlled by women were more likely to report that they had never paid a bribe, and that women tended to have a lower stated tolerance for corruption.

These two foundational studies demonstrated that gender differences in corrupt actions and attitudes towards corruption exist. For more information related to gender and the work of the World Bank, see here.

More recent research, however, suggests that findings on the correlations between gender and corruption are mixed, and that the contexts in which women face corruption and the likelihood that they may engage in it are critical. Other scholars have also analysed correlations between higher female representation in government and reduced levels of corruption. Their studies question findings which suggest that gendered differences in corruption are innate rather than a product of the person's environment. Sung (2003, p. 718), for example, found that "although female participation in government may be correlated to lower levels of corruption under some circumstances, this association loses significance when the effects of constitutional liberalism are appropriately controlled for". Rheinbay and Chêne (2016, p. 4) pointed out that there are many variables which must be considered and that "correlation does not imply causation". Even Swamy and others (2001, p. 26) acknowledged that the gender differences they observed were possibly attributable to factors such as "socialization, or to differences in access to networks of corruption, or in knowledge of how to engage in corrupt practices, or to other factors". Isolating the effect of gender on corruption has proved to be difficult.

To summarize some of the research available on gender and corruption, Boehm (2015) focused on three specific issues:

  • Gender differences in accepting bribes
  • Gender differences in offering bribes
  • Gender differences in attitudes toward corruption

Boehm summarized his research findings on gender differences in corrupt behaviour and attitudes in Table 1.

Table 1: Summary of Boehm's research findings

Attitude towards corruption

Accepting bribes

Offering bribes

Survey responses suggest a lower tolerance of women towards corrupt behaviours.

However, it seems that this is true only in democracies, but not in autocracies and environments where corruption is endemic and widely tolerated.

With some exceptions that seem to respond to contextual factors, there is no significant difference between women and men.

Men are more likely to offer bribes.

Women behave more opportunistically: they may accept the bribe, but do not reciprocate with a corrupt favour.

The value of bribes offered by men tends to be higher.

Women are less likely to accept bribes than men when there is a perceived risk of sanctions.

Both men and women offer higher bribes to men than to women.

Source: Boehm (2015, p. 3).

For a further discussion of the ways in which environment, psychology and situation can influence ethical decision-making, see Module 6 and Module 8 of the E4J University Module Series on Integrity and Ethics.

Next: Theories explaining the gender-corruption nexus
Back to top