Source: PENNSYLVANIA STATE UNIVERSITY submitted to
INVESTIGATING MECHANISMS BETWEEN TIME SCARCITY AND HEALTHFULNESS OF FOOD CHOICES – EVIDENCE FROM SCANNER DATA AND LAB EXPERIMENTS
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
EXTENDED
Funding Source
Reporting Frequency
Annual
Accession No.
1025118
Grant No.
2021-67023-33823
Project No.
PENW-2020-06218
Proposal No.
2020-06218
Multistate No.
(N/A)
Program Code
A1641
Project Start Date
Jan 1, 2021
Project End Date
Dec 31, 2024
Grant Year
2021
Project Director
Jaenicke, E. C.
Recipient Organization
PENNSYLVANIA STATE UNIVERSITY
408 Old Main
UNIVERSITY PARK,PA 16802-1505
Performing Department
Agricultural Economics & Rural
Non Technical Summary
Consumers' time pressures and preferences for convenience products play important roles in Americans' diet-quality choices, which can be partially summarized by two stylized facts: (1) Time pressures and time constraints are strongly linked to poorer diet quality; and (2) Convenience foods are more expensive and less healthy than non-convenience foods. While the empirical links between time and diet quality and between convenience attributes and health are both well established, previous research has not documented the behavioral mechanisms that govern unhealthy food choices under time scarcity nor the policy implications of this behavior.By analyzing data from household-level food purchases and lab-based experiments, the proposed research seeks both to identify the willingness-to-pay differences for health and convenience product attributes across a variety of healthy/convenience products categories and to investigate the behavioral mechanisms that affect consumers' food choices under time scarcity. Our project has four specific research objectives:Investigate the impact of time scarcity on heterogeneous households' food choices and the trade-off between time and healthfulness through reduced-form and structural analysis of consumer-level food-purchase data.Identify potential channels or mechanisms through which time scarcity may affect the healthfulness of households' food choices, and test these mechanisms using lab-based economic behavior experiments.Assess consumer's willingness to pay for time convenience, healthfulness, and their trade-offs, and compare results obtained from both experimental and food-purchase data.For alternative behavioral mechanisms that link time scarcity to food choices, identify and evaluate the effectiveness of various policies aimed at improving the healthfulness of food choices.
Animal Health Component
0%
Research Effort Categories
Basic
25%
Applied
75%
Developmental
0%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
60762993010100%
Knowledge Area
607 - Consumer Economics;

Subject Of Investigation
6299 - Marketing, general/other;

Field Of Science
3010 - Economics;
Goals / Objectives
The long-term goal of the proposed project is to provide research that supports the promotion of healthy food choices and to ultimately improve human health. The overall objectives of the proposed research are to investigate the behavior foundation of the role of time scarcity and food choices and to explore how particular mechanisms, neoclassical or behavioral, may influence the effectiveness of potential food policies. By analyzing data from food purchases and lab-based experiments, the proposed research seeks both to identify the WTP differences for health and convenience product attributes across a variety of healthy/convenience product categories and to investigate the behavioral mechanisms that affect consumers' food choices under time scarcity. Our long-term goal and overall objectives are supported by the following four project-specific research objectives:1. Investigate the impact of time scarcity on heterogeneous households' food choices and the trade-off between time and healthfulness through reduced-form and structural analysis of store- and consumer-level food-purchase data.2. Identify potential channels or mechanisms through which time scarcity may affect the healthfulness of households' food choices, and test these mechanisms using lab-based economic behavior experiments.3. Assess consumer's willingness to pay for time convenience, healthfulness, and their trade-offs, and compare results obtained from both experimental and food-purchase data.4. Identify and evaluate the effectiveness of various policies aimed at improving the healthfulness of food choices under alternative mechanisms that link time scarcity to food choices.
Project Methods
Part 1: Scanner dataThe proposed project will utilize a combination of several large datasets, including (1) IRI InfoScan data, (2) IRI Consumer Panel data, and (3) IRI MedProfiler consumer health data.The IRI InfoScan data cover dollar sales, volume sales, and prices for food products in major national and regional chain food stores. Also, it provides detailed information on product characteristics (e.g. brand names, container sizes, package sizes, etc.), marketing (e.g. price promotions and in-store displays), location, and time of sales.The IRI consumer panel data include over 120,000 households, with 46 to 52 percent of the households providing sufficient purchase data to be included in the static panel for analyses. The data include key demographic variables including household size, age of head of household, income, ethnicity, race, children, and related characteristics. In addition, we will link the IRI MedProfiler data with the IRI consumer panel data, which provide a more complete picture of consumer health and wellness.We first explore the data to corroborate or establish some stylized facts about convenience foods, and the trade-offs between convenience and healthfulness. First, we compare the healthfulness of convenience and non-convenience food. Specifically, we define convenience foods purchased in stores as foods that are Ready to Eat, i.e., foods that can be consumed as is, or Ready to Heat, i.e., foods that can be quickly heated in a microwave. We investigate whether convenience foods are less healthy in terms of containing higher levels of added sodium, added sugar, saturated fat, and cholesterol, which are bounded in recommended consumption by Dietary Guidelines for Americans. Also, we will use the Guiding Stars program to quantify and compare the comprehensive healthfulness of convenience and non-convenience food items.Second, we analyze whether households with tight time constraints purchase more convenience and less healthful foods as a whole. The overall healthfulness of households' food purchases will be measured by the Healthy Eating Index (HEI) and Nutrient Density Score (NDS). The HEI, which varies is used to measure how closely a diet adheres to the recommendations of DGA.Households' time constraints are proxied by (1) employment status, i.e., whether female household head is employed and her hours of working; and (2) childcare, i.e., the number of children, presence of young children (age<6) in particular and if the family has a single parent. The presence of children likely affects households' food choices not just through time constraints but also via other channels such as health concerns.Third, we focus on specific product categories that vary both in preparation time (slow/non-convenience vs. quick/convenience) and levels of healthfulness. One possible example is the soup category, which has both quick versions (e.g. can soup)) and slow versions (e.g. soup mix), and also varies in its levels of healthfulness (soups with different amounts of added sodium and saturated fat). We will analyze how time constraints, education, income, and other important characteristics are related to households' choices between slow-healthy, slow-unhealthy, quick-healthy, and quick-unhealthy versions of the same food category.Structural modeling is necessary to account for a large number of product choices and allow for flexible substitution patterns and realistic distribution of consumer preferences among heterogeneous consumers, Product attributes reflecting healthfulness and convenience of food preparation will specifically be included in the model. We will then use the estimated parameters to conduct simulations to evaluate the impact of time scarcity on the healthfulness of consumers' food choices.The estimates from our demand models will be used to calculate consumers' willingness to pay (WTP) for convenience and healthfulness. The ratio of the Quick and Healthfulness coefficients to the price coefficient is a measure of the amount a consumer is willing to pay for these features of soup and eventually other food products.Based on the initial case study for soup (plus additional product categories, eventually), estimated parameters will be used to simulate the food purchases with different convenience and healthfulness features when alternative policies are implemented.Part 2: Lab experimentsIn the project's second phase, we plan to conduct a multipart laboratory experiment to study the impact of time scarcity on the healthfulness of food choices as well as the underlying behavioral mechanisms. The empirical association between time scarcity and unhealthfulness of diet might be mediated by time scarcity at the moment of decision and/or at the moment of eating. Furthermore, time scarcity can be encountered when preparing food at home. In the laboratory experiment, we investigate time scarcity in these three forms and the behavioral mechanisms that might cause a change in the healthfulness of food choices.We plan to invite approximately 300 subjects to participate in the proposed lab experiment. We will invite them to our laboratory facilities at Penn State University in small groups of 10 - 15 people. The lab experiment will consist of three parts: 1) Willingness to pay (WTP) for convenience and healthfulness, 2) Lunch selection, and 3) Survey questionnaire.In the first part, we study the link between time scarcity and healthfulness of food purchases, and we focus on preparation time as the form of time scarcity in this experimental design. Based on an auction mechanism, we elicit participants' WTP for foods' healthfulness in foods that require different preparation times. Under neo-classical assumptions, the WTP for the health-related attributes of foods should not differ for different preparation times. As in the scanner data portion of the project, we are interested in whether or not complementarities exist between health and convenience. If participants' WTP for healthfulness is found to be lower for foods that require shorter preparation time (quick) than slow food, this result would indicate a complementarity between quick and unhealthy food. This result would support the hypothesis of a behavioral response to time scarcity and a preference for energy-dense "comfort foods" when time is scarce.In the second part of the experiment, the lunch selection, we investigate the relationship between time scarcity and healthfulness of food consumption. We invite participants to have lunch in our facilities and randomly assign the time available to consume their meals. If subjects with less time to consume the meal select and consume more unhealthy options, this result would support the hypothesis that time scarcity leads to a behavioral (i.e., physiological or emotional) response and a preference for unhealthy.Alternatively, the impact of time on food choices might be mediated by the time available to make the food selection. In both parts of the experiment, we randomly vary both the time available to state the WTP and to select the preferred lunch option. If participants that are randomly allocated less time have a higher WTP for unhealthy products and make less healthy meal selections, this result would support the hypothesis that time scarcity impacts the healthfulness of food choices by creating time-stress and bounding their rationality, i.e, preventing individuals from thoughtfully and carefully evaluating their choices. It is possible that with little time to decide, individuals could dismiss the value of the healthy attribute, thus leading to the selection of a less healthy food option.

Progress 01/01/22 to 12/31/22

Outputs
Target Audience:No changes. The target audience for this project is (a) academic peers in agricultural, food, and applied economics, (b) food and health policymakers, and (c) food industry analysts. Changes/Problems:While the pandemic delayed the start of the project's in-person experimental portion forced changes in our institution's face-to-face research protocols involving food, IRB-human subject approval is place. What opportunities for training and professional development has the project provided?One Ph.D. student, Sihyun Park, helped with Goals 1 and 2, and is preparing for her comprehensive exam/PhD proposal. How have the results been disseminated to communities of interest? Nothing Reported What do you plan to do during the next reporting period to accomplish the goals?Goal 1: In the next reporting period, the study team plans to (a) refine existing econometric models to allow for household heterogeneity. This refinement will allow us to see if our preliminary findings vary for low income or wealthy households, low-BMI or high-BMI households, or for other varying household characteristics. Goal 2: In the next reporting period, the study team will finalize the experimental design and instructions, along with the accompanying survey, register the experiment with the American Economics Association, and pilot the experiment. If all goes well with the pilot, the study team will fully implement the experiment during the next reporting period.

Impacts
What was accomplished under these goals? During the reporting period, the study team made progress on the to main parts of the project, that is, (i) the analysis of convenience and healthfulness product attributes suing the IRI consumer panel and retailer scanner data, and (ii) the design of the lab experiment investigating behavioral mechanisms for consumer preferences for convenience and health. Goal 1: Using the scanner data, the project team added convenience and health attributes to the IRI data for oatmeal products and canned soup projects, and then estimated consumer demand models for oatmeal and canned soup that included these attributes. More specifically, healthfulness was quantified using third-party "Guiding Stars" methodology and convenience was quantified using cooking times - either in minutes or using fast or slow categories. With these attributes coded, the study team next estimated models of consumer demand. Preliminary results suggest that households value healthfulness and convenience positively, but that their interaction has a negative impact. The interaction means that, for example, that increased healthfulness is valued less for products that have increased convenience. These preliminary results, so far, hold for both oatmeal and canned soup. Goal 2: The study team has begun designing the lab experiment and accompanying survey. In short, the lab experiment has two parts, with Part 1 allowing participants to choose a variety of snack food products across the healthy-unhealthy spectrum while varying at least two aspects of time pressure - that is, shortness of time and stress. In Part 2, participants take part in a "auction" for food products after being exposed to alternative treatments that also involve time stress. Finally, this part of the project is accompanied by a survey of participant attributes and attitudes. The study team has draft versions of the experimental design and instructions, as well as the survey.

Publications


    Progress 01/01/21 to 12/31/21

    Outputs
    Target Audience: Nothing Reported Changes/Problems:During this reporting period, the study team recruited and hired a graduate student to work on the project. The pandemic delayed the start of the project's in-person experimental portion and our institution changed a number of face-to-face research protocols, especially those where the experiments involve food and eating. The study team also finalized IRB-Human Subjects approval near the end of the reporting period. What opportunities for training and professional development has the project provided?As noted above, the study recruited and hired a PhD student for this project. The student started working on this project in September 2021 and meets with the faculty-led study team once a week. How have the results been disseminated to communities of interest? Nothing Reported What do you plan to do during the next reporting period to accomplish the goals?Goal 1: In the next reporting period, the study team plans to collect data on additional products and begin estimating discrete choice models of consumer demand. Ultimately, these estimated models will be used to find the value of both a convenience-based product attribute (e.g., cooking time) and a healthfulness product attribute (e.g., the number of "Guiding Stars). Goal 2: In the next reporting period, the study team plans to finalize design of in-person experiments to investigate how and why participants make choices over convenience and health under different treatments related to time pressure.

    Impacts
    What was accomplished under these goals? For Goal 1: During the reporting period, the study team gained access to IRI and MedProfiler data on the secure NORC data enclave. After access was obtained, the study team began analyzing IRI data for several products with both convenience and health product attributes. Preliminary products include canned soup and oatmeal. For Goal 2: The study team also secured IRB-Human subjects approval, which was a more complicated process because of new Covid-19 restrictions.

    Publications