You walk through a shopping mall and spot a pair of expensive sunglasses you have long admired on sale. You debate the pros and cons of splurging and decide not to buy them, but turning a corner, you see that your favorite chocolate shop has a new product on its shelves. It's pricey for chocolate but within your budget, so you buy a box.

Or, you have the means to afford the sunglasses but don't really need another pair. However, 15 per cent of the price will go to charity so you decide to buy them. You have the same encounter with the chocolate shop and again you buy a box.

These scenarios illustrate how justification, based either on prior shopping restraint or the purchase of something seen as virtuous, can lead people to subsequently make an indulgent purchase. These arguments are presented in research by Anirban Mukhopadhyay of HKUST and Gita Venkataramani Johar.

"Our main point is that a salient memory or restraint in the face of such temptation, sanctions consumers to reward themselves when a subsequent temptation presents itself," they say.

They conduct several experiments that support this idea. Two of these demonstrate that people who show restraint in a first shopping incident are about twice as likely to indulge when presented with a tempting product later on, as those who do not show restraint. However, this only holds if the prior purchase is made salient - in this case, by asking participants to recall why they had shown prior restraint.

Interestingly, when the authors ask them why they made the second, indulgent purchase, participants with heightened salience are much more likely to agree that "I had refrained from buying the books [in the first incident], so I deserve a little treat".

This justification supports the authors' view that people will indulge if they think they have a good reason to do so. A third experiment tests this idea by showing that if people give in to temptation at the first instance but feel good about it because it involves a charitable donation, they will also subsequently indulge.

In this case, participants are presented with a chocolate bar without warning. Some are told that if they purchase it the money will be donated to charity, some are told nothing. They then do unrelated tests and are given a choice at the end whether to have chocolate cake or a fruit salad.

"As expected, in the no-donation condition, respondents preferred cake more when they had bought the chocolate bar than when they had. But the results reversed in the donation condition: those who had not bought preferred the cake less than those who had bought."

"This suggests that it is not the act of buying or not buying that causes indulgence; rather, it is the justification afforded by that act."

The authors say the results have implications for marketers and those who track consumer purchases over time in terms of how and when to present tempting offers.

"Based on our findings, managers would be advised to sequence offers such that indulgences are offered following a no-buy decision. Managers could position an indulgent product by making the customer think about their previous restraint, thereby bracketing the two together," they say.