Complex Pricing and Financial Stress
How to cite
Tjernstrom, Emilia; Michler, Jeffrey. 2018. "Complex Pricing and Financial Stress." AEA RCT Registry. AEARCTR-0002904. https://doi.org/10.1257/rct.2904-1.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
A vast literature in economics has considered the question of how firms set prices, and a growing literature has demonstrated how real people differ systematically from theoretical consumers. In this project, we ask how real consumers in an experimental setting respond to firms' prices. We focus on a specific pricing strategy, bundling of multiple goods, which has been shown to be theoretically optimal in a wide variety of settings (Chu et al., 2011; Eckalbar, 2010) but that may result in long, complex menus of prices that are difficult for consumers to digest. We aim to study (1) whether these complex menus drive consumers to make sub-optimal choices, and (2) how those real choices might change optimal firm behavior. We also test the impacts of two firm strategies that affect the complexity of the choice problem that consumers face: unexpected price shocks and information costs. We will quantify how these strategies may cause consumers to differentially deviate from an optimal choice, and test whether these may have lingering spillover effects on future decisions. We study these questions using a combination of simulated firm behavior with results from a lab experiment with students in the BRITE Lab at the University of Wisconsin-Madison.
Intervention
We define uniform ranges of consumer valuations for a set of experimental goods. Using these distributions, we simulate a large number of markets in which monopolist firms offer subsets of the experimental goods using one of four pricing strategies, and simulated consumers choose goods or bundles based on randomly drawn valuations. We use these simulated-optimal prices for every pricing strategy and subset of experimental goods as inputs for the lab experiment, in which participants are asked to make a series of purchasing decisions given a private draw of reservation values. The experiment interface includes prices, personal valuations, and allows participants to click to view the surplus calculation of a given selection.
We focus on four pricing strategies:
1. Component Pricing (CP): firms offer one price per good and no bundle discounts;
2. Pure Bundling (PB): firms offer a single price for a bundle of all goods on sale;
3. Mixed Bundling (MB): firms offer a price for every good and every possible combinations of those goods;
4. Bundle-Size Pricing (BSP): firms offer one price per bundle-size, regardless of which goods are included in the bundle.
The experimental consumers will face different prices and different numbers of goods, and we will analyze how these affect the optimality of consumer choices.
We additionally include two random treatments:
1. Price shocks: firms inform randomly selected participants in randomly-selected rounds that the firm is targeting them with higher prices due to their personal characteristics
2. Information costs: firms inform randomly selected participants in groups of random rounds that they have reviewed their costs and will now charge for revealing surplus information. Information costs can either occur early or late, and these are randomized as separate treatments.
We will use the results of the experiment to characterize consumer behavior. In particular, we want to understand how the complexity of a decision, determined by pricing strategy, number of goods, and the two treatments, may cause consumers to deviate from optimum.
Experimental design
We conduct our lab experiment with a target sample size of 1200. Participants are randomly assigned private reservation values for 6 fictitious goods at the start of the game. Participants complete 46 rounds of a game in which they are offered a random subset of 2-5 of goods, using one of four randomly selected pricing strategies: component pricing (CP), pure bundling (PB), mixed bundling (MB), and bundle-size pricing (BSP).
The treatments are randomly assigned at the individual level. The design is fully factorial, with 6 possible assignments A-F (as per below):
Price shock No price shock
Information shock early A B
Information shock late C D
No information shock E F
We measure participants' final choices, as well as any interim clicks and the time spent on each round. Between each round, we also ask participants how they are feeling after each round using a list of emojis. The list is designed to have a balance of positive and negative, and a balance of engaged and disengaged, and a neutral choice. The order of the list is randomized each time. With this, we hope to directly link elicited emotional responses to game behavior.
Design details
Cited by
No citation data available for this registration yet.