567  2006 by JOURNAL OF CONSUMER RESEARCH, Inc. ● Vol. 32 ● March 2006 All rights reserved. 0093-5301/2006/3204-0010$10.00 Habitual Behavior in American Eating Patterns: The Role of Meal Occasions ADWAIT KHARE J. JEFFREY INMAN* Based on literatures in cognitive resource conservation and contextual cue con- sistency, we study two types of habits—carryover and baseline—in the consump- tion of food nutrients. Carryover habit obtains when the level of a nutrient consumed in preceding meals influences its consumption in the current meal. Baseline habit obtains when a nutrient’s consumption systematically differs across meals. We test our hypotheses via a hierarchical linear model using a food consumption diary panel. Findings support our carryover habit and baseline habit dichotomy, as well as our predictions that carryover habit is stronger at breakfast and that within-meal carryover effects are stronger than across-meal carryover effects. P eople have neither the mental resources nor the time to constantly perceive, evaluate, and then act with respect to every aspect of life (Simon 1982). Habits—behaviors per- formed frequently and consistently in stable contexts—are one solution to this problem. Habits have been described as learned and scripted behaviors (Aarts, Verplanken, and van Knippenberg 1998) toward which consumers have been con- ditioned (Bargh and Ferguson 2000). By requiring fewer cog- nitive resources for their operation, habits help consumers conserve these limited resources and free them for other tasks (Wood, Quinn, and Kashy 2002). For example, while eating breakfast, many of us are in a hurry and are thinking about activities during the day ahead. Food consumption habits can free a person from exerting effort in repetitive decisions. In this research, we study the nature of habits with respect to the consumption of food nutrients. Food consumption is frequent (at least once every day), consistent (everyday), and done in stable contexts (at home for most consumers). Since the link between past and future behaviors is stronger when actions are performed more frequently and in stable contexts (Ouellette and Wood 1998), habits are likely to be common in food consumption behavior. This research makes several important contributions. Our first contribution is the conceptualization and demonstration of different magnitudes of carryover habit, which describes the effect of past behaviors on current behavior. Specifically, *Adwait Khare is assistant professor of marketing, C. T. Bauer College of Business, University of Houston, Houston, TX 77204 (akhare@uh.edu). J. Jeffrey Inman is Albert Wesley Frey Professor of Marketing, Joseph M. Katz Graduate School of Business, University of Pittsburgh, Pittsburgh, PA 15260 (jinman@katz.pitt.edu). The authors thank the editor, the review team, Larry Feick, John Hulland, Ajay Kalra, and Vikas Mittal for their detailed and insightful comments on previous drafts. The authors are grate- ful to the NPD Group for providing the data used in this research. we find that across-meal carryover effects (e.g., the impact of today’s breakfast and lunch on today’s dinner) are weaker than the within-meal carryover effect (e.g., the impact of yesterday’s dinner on today’s dinner). Our second contri- bution is the conceptualization and demonstration of base- line habit. We argue that the baseline nutritional qualities of the three daily meals systematically differ from one an- other. Baseline habit provides a new way of conceptualizing habits, as it addresses the contextual specificity of repetitive behaviors such as food consumption behavior. While vari- ation in food consumption behavior is to be expected given the variety of forces that influence eating behavior, baseline habit demonstrates that there are stable patterns to this var- iation. Our third contribution is the demonstration of dif- ferences in carryover habit based on meal occasion (break- fast, lunch, or dinner). Specifically, we argue that contextual cues are more stable at breakfast and, therefore, that the difference between within-meal and across-meal carryover should be greater at breakfast than at lunch or dinner. The results are supportive of this thesis. In the next section, we elaborate on the conceptual un- derpinnings of carryover and baseline habits and develop our hypotheses. Following that, the food consumption data and the hierarchical linear modeling (HLM) methodology we use to test our hypotheses are discussed. We then present the results and discuss implications for research on consumer habits and for substantive applications (e.g., nutritional guide- lines and weight management). CONCEPTUAL BACKGROUND Baseline Habit Baseline habit results in systematic differences in nutri- ents’ intake levels across meals. Research in the food con-