Sunday, December 29, 2013

Does Dynamic Pricing Have a Speed Limit? Should It?

On the heels of the big Uber flap over its dynamic pricing model, and Matt Yglesias exploring alternatives to explicit dynamic pricing schemes in the restaurant industry, there's been a lot of discussion over the variability of prices lately. The reasons are pretty clear: technology and the internet have crashed the costs associated with setting up transactions down to essentially zero. A corollary of this is that with an algorithm (or the click of a button) you can change prices constantly. Input commodities have had rapidly-changing prices for a long time, but front-end dynamic pricing has been limited in the customer service domain (hotels, lobster shacks, etc.) Clearly we're seeing the conflict that arises when people's norms and expectations don't keep pace with technological change.

Dynamic pricing seems to be mostly a good thing. It uses resources more efficiently (along the intensive margin), allows more price discrimination (charging people what they're willing to pay), and incentivizes new supply during peak times. But in some areas there are probably limits to the social gains to be had from ever-faster price changes. The best example of this is high-speed trading in financial markets, which now operate at speeds far exceeding what their supposed capital-allocation function requires.

As parts of the economy become more "service-and-flow"-oriented (also called the "sharing economy") and less ownership-based, the moral implications of constantly-shifting prices become apparent. If few people own cars and simply order the service of road transportation (from self-driving cars, or cheap ride-sharing programs), everyone becomes very exposed the uncertainty of changing prices. Certain groups (like those with money!) will be better able to cope with the added uncertainty. If you can't say for certain how much your parking meter is going to cost before you leave the house, maybe you'll just stay home.

Perhaps everyone will simply adapt and routinize price-checking in the morning alongside their coffee and newspaper, but for people who's mental bandwidth is already stretched this could be a problem. Lifestyle complexity is regressive. From this perspective, certain forms of regulation can be justified on the grounds of limiting the adoption of dynamic pricing. I've tended to be a big supporter of eliminating building and rental/housing restrictions, but imagine what a truly efficient housing sector would mean in sellers-market conditions: would renters have to pay different amounts every month? Every week?

We're already seeing business trying to reduce the stickiness of their labor costs by rebalancing to more part-time and temporary workers. While the macroeconomic effects of this trend might be nice (for GDP growth at least, not so much for unemployment), more and more dynamic pricing in key sectors could also have worrying pro-cyclical implications. The effects of demand shortfalls could more easily (and more quickly) ripple through previously isolated areas of the economy. How these competing effects of faster price changes actually compare in the data is a underexplored topic for researchers.

No comments:

Post a Comment