Experimental economics: Price Matching Guarantees (2)

Hi again everybody (so far it seems no one has left any comments... but let's move on)
This is the update regarding Price Matching Guarantees (PMGs), where one seller promises to match (note: match, not beat) its rivals prices, as seen in supermarkets or bookstores. I will recap the theory again. When a seller says that he/she will match its rivals prices on a product, it seems like such cut-throat competition, doesn't it? The truth is that it is anti-competitive. Think about it this way. If I sell my product for $70 (assume a price range of 1 to 100), then if you sell yours for $65, you will service the entire market demand for that good. However, if I promise to match your price of $65, then we both split the market. Since that's the case, we might as well tacitly collude (in layman's terms, cooperate non-formally) and set our price at $100 and get even more profits. There's no point in you lowering your price to $60 if you know I'm going to match it - and hey, that means lower profits for the both of us! (Assume that the number of buyers and products on the market remain constant). So, we might as well keep a high price. And that is the anticompetitive nature of PMGs.

There are many assumptions that come along with the theory of course (which was first noted by Salop, 1986). I will skip them, but the main point to note is that we are assuming that it doesn't cost buyers anything to invoke the guarantee. BUT, other authors have noted that if buyers have to incur some sort of 'hassle cost' - i.e. such as fill in forms, go back to the shop and ask for the guarantee slip, etc, which cost money - then buyers would rather just buy from the seller that immediately offers the lowest price on the market, regardless of if its competitor chooses to match its price. In such a case, the game unravels and the predicted competitive equilibrium is equal to marginal cost (in this case, the lowest price of $1). True enough, the experimental data confirms this convergence towards the equilibrium price, although full convergence does not necessarily happen.

See: Dugar and Sorensen (2006) for example (sorry I can't link the article here due to copyright)