[nanis]

> the "Evidence from india" part is important.

Why? The theory is independent of racial or national origin:

    Economic theory predicts that civil servants often shirk
    responsibility or take bribes because it is difficult for
    central governments and citizens to monitor and subsequently
    punish these bad behaviors ( e.g., Banerjee 1997, Shleifer and
    Vishny 1993, Di Tella and Schargrodsky 2004, and Olken and
    Pande 2012) . This implies that variation in the ability to
    monitor or incentivize civil servants may drive the observed
    differences in corruption across countries, across agencies
    within a country, or even across the types of tasks for which
    public servants are responsible. However, not all civil
    servants engage in the same level of corrupt behavior, even in
    the same position or role. Besley (2005) and Prendergast (
    2007) posit that this may be potentially due to different
    government workers having different preferences over engaging
    in corruption. As such, it follows that the types of
    individuals that select into government may help explain
    variation in corruption levels.

[frobozz]

It is independent of racial and national origin of the individual concerned, but not independent of the nation of the civil service being examined.

> This implies that variation in the ability to monitor or incentivize civil servants may drive the observed differences in corruption across countries across agencies within a country, or even across the types of tasks for which public servants are responsible.

> As such, it follows that the types of individuals that select into government may help explain variation in corruption levels.

As I stated in my comment on this, where bribery of public servants is the norm (with this statement comes an implied "and typically goes unpunished", otherwise it wouldn't be the norm), one might expect corrupt individuals to be attracted to those roles where they can pocket the most bribes.

In countries (or agencies within a country) where such bribery is rare, and dealt with seriously if it does occur, what would be the attraction for a corrupt individual?

It could well be true that test-cheaters are attracted to the public sector, regardless of how corrupt that particular public sector is, and only exercise their naughty side in countries/agencies/positions where they can get away with it, but supporting evidence for that assertion (against the assertion that test-cheaters are attracted to positions where they can exercise their naughty side) would only be found in places where they are less likely to get away with it.