RESEARCH

Working Papers

Shrinkflation: What Happens When Package Sizes Change? (Job Market Paper)

(With Pedro Pires) [Draft]

Package size is a central characteristic of consumer goods, and changes in size have implications for demand, welfare, policy, and firm behavior. This paper first develops a method to identify size changes at scale. We apply it to retail scanner data and show that size reductions are considerably more common than increases. Next, we estimate dynamic demand responses to size changes with an event study. Following size increases, demand rises immediately and remains stable. In contrast, after size decreases, demand initially falls modestly but continues to decline over time. An incentivized online shopping experiment shows that individuals deviate from optimal behavior by responding less to size than to price, and provides evidence of attention as a mechanism. We explain the empirical results with a dynamic demand model that allows for inattention, paired with a supply model in which firms strategically advertise size increases but not reductions. We then derive a sufficient statistic for consumer surplus losses from inattention to sizes and quantify it using estimates from the retail scanner data and the experiment. Following a median size reduction, short-run losses from inattention are equivalent to the deadweight loss from a 6% tax. We also evaluate the policy interventions included in the experiment. Displaying unit prices and making sizes more prominent bring demand responses closer to the optimal benchmark, while size change labels induce overreaction to sizes. Our welfare analyses show that the first two interventions substantially reduce consumer surplus losses from inattention, while the third yields only modest improvements.

Fake News: What Is Its Impact? Can We Do Something About It?

[Draft]

The rise of fake news led to widespread public concerns based on claims that it affects behaviors. Nevertheless, the evidence of the impact of fake news on behaviors is scarce and inconclusive, and there is no consensus on how to deal with the potential negative effects. In this paper we conduct a survey experiment to evaluate the impact of a fake news article concerning environmental activism. Additionally, the effect of a fact-checking article correcting the fake news story is also evaluated. The measured outcomes include beliefs and pro-environmental preferences and behaviors. The treated participants were randomly assigned to either read an extract of the fake news article or of the correction article. The remaining participants constitute the control group. We find clear impacts on beliefs: the fake news article increases misperceptions while the correction reduces them. However, our results show small to null effects of both treatments on pro-environmental preferences and behaviors.

Selected Work in Progress

The Role of Price History and Expectations in Consumer Choice

(With Alexander Whitefield)

What Determines Support for Labor Movements? An Analysis of a Randomized Digital Campaign

(With Junru Lyu)