Let's start off with some evidence of Giffen goods and rice.
"...show that in each case Giffen behavior is closely associated with poor consumers’ need to maintain subsistence consumption in the face of an increase in the price of a staple commodity. We then present evidence on the existence of Giffen behavior among extremely poor households in two provinces of China. In order to obtain an unbiased estimate of the key price elasticity, we conducted a field experiment in which we randomly subsidized households’ primary dietary staple (rice in Hunan province and wheat flour in Gansu province). Using consumption data gathered before, during and after the intervention, we find strong evidence of Giffen behavior with respect to rice in Hunan province."
This suggests that rice is a Giffen good, remember that a Giffen good is an extreme type of inferior good. The negative income effect of changes in price of a Giffen good is actual stronger than the substitution effect. This leads to its bizarre quality: when the price of a Giffen good rises, consumers actually buy more.
But what has this got to do with prostitution?
Quoting from the Prostitution index...
Prostitution is a unique labour market. Most people find it an extremely undesirable job, but on the high end, it can be quite lucrative and requires few skills (though a fair helping of unequally distributed natural endowments). These factors make the prostitution market exceptionally sensitive to large fluctuations in wealth and expectations.
The most interesting part is that the market is counter cyclical. In bad times more (and more attractive) women enter the market, but they have a higher reservation wage, so they charge more. Less attractive and even cheaper prostitutes may still be available, but for a variety of very good reasons, the customer will not desire the cheapest option, suggesting prostitution services can be classified as a Giffen good.
Does this mean that you can quote prostitutes as being a Giffen good in your Economics papers?