Experimental Design
This online experiment will take place in November-December 2020. Each participant will be asked to complete a 10-15-minutes online survey. We aim to recruit approximately 1,000 respondents from the Mid-Atlantic Region in the U.S., with the help of the survey provider Dynata, LLC.
In the first stage of the study, we randomly assign respondents to three conditions
1. Individual
2. Community
3. Control
In the Individual condition, we ask respondents about the impact of the coronavirus pandemic on their personal life and their personal financial and health concerns, and then ask them to describe the most important ways in which the pandemic has impacted their life. In the Community condition, we ask respondents about the impact of the coronavirus pandemic on their local community and their concerns for their community, and then ask them to describe the most important ways in which the pandemic has impacted their community. Asking participants to think about the impact of Covid-19 places the crisis at the top of the respondents’ mind, creating the context in which they would then consider the following questions and decisions.
In the second stage, we firstly define local food as food that was grown, produced, or processed within 50 miles from the respondents. Participants are then asked about their preferences for local food through a series of choice-based questions meant to elicit a hypothetical price premium. We select this definition of local amongst many others as commonly supported by the literature (Chambers et al., 2007; Groves, 2005; La Trobe, 2001). Moreover, Darby et al. (2008) find that consumers assign the same value to different definitions of “local” and that willingness to pay for a local attribute is independent from other features. Also, in the study we are mainly interested in the difference between treatments and not on the exact definition of local foods.
We measure the price premium for local fruit or vegetables and the price premium for local meat or poultry products with two sets of questions. The questions are adapted from Carpio et al. (2009). We select food categories instead of specific food items to avoid food-specific aspects other than preference over the local attribute (as disliking the food selected, food allergies, religious prohibitions, and others) would drive responders answers (Cranfield et al., 2012). We collect the self-reported price premium for the local food using the “unfolding brackets method”. Participants are asked are asked to indicate what they would buy between local and non-local foods, at varying price premiums for local food. The unfolding brackets mechanism involves an adaptive questionnaire with binary choices presented to participants depending on their prior choices. The first choice of the titration is between non-local and local food products at equal prices. If participants choose the local food, then the next choice will be between non-local food and 50% more expensive local food. The alternatives adjust upwards or downwards to narrow in on the indifference point (accurate to a 5% interval).
We then measure participants’ donations to charitable organisations supporting farmers, farmers markets, or a food relief program. We notify participants that a randomly selected group of participants will receive a $25 bonus. We ask participants, should they be selected for the bonus, whether they want to donate part of the $25 to three charitable organisations or keep it all for themselves. To capture individual preferences for certain charitable organisations, we ask participants to make the allocation decisions between themselves and three charities. We have selected a charity supporting farmers (Farm Aid), one supporting farmers market (Farmers Market Coalition) and one committed to bringing food to people in need (World Central Kitchen).
Respondents are then asked about their sense of community belonging using a question by Carpiano et al. (2011), and to complete the six-item short-form state anxiety inventory developed by Marteau and Bekker (1992). Afterwards, we collect respondents’ shopping habits and their agreement with six attitudinal statements regarding personal motives for purchasing local food. These responses will help capture attitudinal factors that can play a role in explaining preferences for local foods. We select the questions based on their relevance in determining preference for local food (Chinnakonda et al. 2007; Grebitus et al. 2013) andbecause these attitudes are likely to be interact with our treatment in strengthening the impact of a change in price premium due to the priming. In particular, we might expect an interaction between safety concerns and the Individual prime, and an interaction between the support for local economy and the Community prime. We ask participants’ agreement with six sentences about motives for preferring local food using five-point scales ranging from 1 (strongly disagree) to 5 (strongly agree). In particular, we ask about support for the local economy (question 1 and 2), safety (question 3), environmental concerns (question 4), and freshness and taste (questions 5 and 6). We also ask respondents about their personal definition for local food, their self-reported altruism, their frequency of charitable giving, whether they themselves or anyone in their family got infected with COVID 19 or lost their job because of the pandemic, and several demographic variables.