Request for Proposal - “Displacement”

Using consumer panels to estimate displacement of animal products by plant-based analogues

Deadline: September 1, 2024


Rationale

Consumer panel data track the grocery purchases of a cohort of households or individuals over time (see point 3 below for examples). A handful of recent studies have illustrated the promise of consumer panel data for measuring the impact of plant-based analogues on animal product sales. For example, Cuffey (2022) uses an event study to investigate the effect of the introduction of the Beyond Burger, while Slade (2023) uses lagged purchases of almond milk as plausibly exogenous changes to nondairy milk consumption that might explain changes in dairy milk consumption.

Goal

Proposals submitted under this RFP should expand this body of work with innovative methodologies and more diverse products to answer the question: Does the purchase of plant-based analogues in retail settings cause displacements of the corresponding animal products?

Details

Successful proposals will meet some if not all of the following criteria:

  1. Focus explicitly on causal inference by developing a target trial (Hernán, 2016) and exploiting potential natural experiments or other exogenous sources of randomness. These might include the introduction of new products, consumers moving from one area to another, exogenous changes in price, food-related health news, or product promotions.

  2. Suggest creative modeling approaches that take full advantage of the available data. In particular, consumer panels offer a rich set of potential covariates that might be used for counterfactual prediction with machine learning methods like association rule learning (Dieng, 2019). Panel data also offers the opportunity to conduct event studies (Sun and Abraham, 2020), and variation in the amount of plant-based analogue purchased might be used to characterize a ‘dose’-response relationship. Difference-in-differences methods based around product introductions or product labeling changes may provide suitable natural experiments. Synthetic controls and matrix completion methods may also be applied.

  3. Use outcome measures which correspond directly to purchasing behavior rather than self-reports. Potential sources might include Nielsen Marketing Data via Chicago Booth Kilts Marketing Center, Kantar, or Kroger’s analytics firm 84.51°. Where possible, outcome measures should focus on the physical volume or weight of plant-based analogue purchases and resulting animal product displacement, rather than dollars spent or purchase frequency.  Similarly, unit prices collected at the point of sale are preferred to prices derived from expenditure. We are looking for proposals from the researchers that either already have access to such data or are willing to apply; our fund is not able to cover data costs in excess of roughly $15,000.

  4. Examine plant-based analogues not yet widely explored in the literature such as liquid eggs, ready-meals (pre-packaged complete meals), coffee creamer, mayonnaise, margarine/plant-based butter, ice cream, or yogurt. Alternately, innovative work considering plant-based milks and meats is also welcome, especially work that conceptually replicates existing findings with different methods. Consideration should be given to the price differential between plant-based analogues and their counterparts; products approaching price parity are of particular interest.


Before beginning your application, please review our FAQs and policies.