Dinner preparation in the modern United States

The Authors

Margaret E. Beck, Center of Everyday Lives of Families, University of California, Los Angeles, California, USA

Acknowledgements

This study is part of an interdisciplinary, collaborative research endeavor conducted by members of the UCLA Center on Everyday Lives of Families (CELF), under the direction of Elinor Ochs. CELF is generously supported by the Alfred P. Sloan Foundation program on the Workplace, Workforce, and Working Families, headed by Kathleen Christensen. The author gratefully acknowledges Elinor Ochs and CELF faculty, staff, and graduate students for their fieldwork efforts, comments, and other assistance. Thanks also go to Carolina Izquierdo (CELF) and to two anonymous reviewers for their helpful comments on this paper. A major debt is owed to the working families who participated in this study for opening their homes and sharing their lives.

Abstract

Purpose – The purpose of this paper is to examine the family dinner in Los Angeles County, California, focusing on the role of commercial foods and the time invested in food preparation. Popular media emphasize the increasing use of processed commercial foods in the USA.

Design/methodology/approach – A total of 64 dinner preparation and consumption events were videotaped and observed (32 families, two weeknights each). Observations determined the source of food served (restaurant, take-out, or home-cooked), the ingredients and dishes in each meal prepared at home, and the time required to prepare it.

Findings – The findings in this paper showed that, even when prepared at home, most evening meals included processed commercial foods in at least moderate amounts. Home-cooked meals required an average of 34 minutes' “hands-on time” and 52 minutes' “total time” to prepare. Heavy use of commercial foods saved, on average, ten to 12 minutes, hands-on time but did not reduce total preparation time. Commercial foods require more limited cooking skills and permit more complex dishes or meals to be prepared within a given time-frame than do raw ingredients. They may also reduce time investment at stages other than meal preparation, such as shopping.

Originality/value – This paper provides a rare glimpse of food preparation and meal consumption behavior on the family level. Most reports on US food habits are based on survey and purchasing data, rather than direct observation of household activities as used here.

Article Type:

Research paper

Keyword(s):

United States of America; Convenience foods; Family.

Journal:

British Food Journal

Volume:

109

Number:

Year:

2007

pp:

531-547

Copyright ©

Emerald Group Publishing Limited

ISSN:

0007-070X

Introduction

The North American diet, as described in the popular media, is one dominated by commercial processed foods. These modern Americans “eat breakfast in their cars, lunch at their desks and chicken from a bucket” (Scrivani, 2005). People in the USA reportedly prepare fewer meals at home than in the past few generations (Visser and Stark, 1989; Scrivani, 2005; Scattergood, 2006; Severson and Moskin, 2006) and often fail to adequately shop and plan beforehand when they do cook (Chatzky, 2006). Families who eat a home-cooked dinner together are presented not as the norm, but as models for emulation (Calta, 2005; Pisano, 2006), sustaining family and cultural traditions, cooking skills, and physical health.

The book Fast Food Nation (Schlosser, 2001) and Morgan Spurlock's 2004 documentary “Super Size Me” captured the US public's attention by describing serious physical, social, and environmental effects of heavily processed foods, and the US Government has evaluated some of these problems (McGinnis et al., 2006). Although the disadvantages have been widely publicized, pre-prepared or partially prepared foods (including heavily processed foods) are still frequently used, arguably to reduce food preparation time when the opportunity cost of time rises (Florkowski et al., 2000). Food market data indicate that in the US, these foods are purchased in ever-increasing amounts (Sloan, 2006).

Any critical assessments or plans to improve diet and nutrition should consider what people actually do on an individual or family level, and how dietary recommendations may fit or conflict with established patterns of behavior. We know surprisingly little about these patterns, in a quantitative sense, on a family or individual level today. Large-scale anthropological research on food in the USA began in the 1940s with the Committee on Food Habits, inspired by nutritional problems observed during the Depression and oriented toward improving American diets (Levenstein, 1993). Marketers now conduct most of the detailed research on meal composition, preparation methods, and time investment, usually through surveys rather than observation (Sloan, 2006). Attempts to link questionnaire data with material evidence, such as household garbage, reveal that people often misreport what they purchase or consume (Rathje and Murphy, 1992, pp. 66-75; see, also, Rathje, 1979, 1984a, b; Ritenbaugh and Harrison, 1984).

This paper describes weeknight dinner preparation in Los Angeles County, California, using direct observations of 32 families. Dinner is defined here as the primary family meal in the evening, usually served by these Los Angeles families around 6 or 7pm. Routine meals on weeknights, such as those described in this paper, make up most of a family's dinnertime experiences. After the basic meal structure is outlined, the focus shifts to two aspects of meal preparation: the role of commercial foods and the amount of time invested.

Commercial foods, as the term is used here, are either purchased as ready-to-eat or prepared by the home cook entirely according to package directions. “Commercial foods” are similar to “convenience foods”, formally defined elsewhere as “fully or partially prepared foods in which significant preparation time, culinary skills, or energy inputs have been transferred from the home kitchen to the food processor and distributor” (Capps et al., 1985). Time saving is not considered in this definition of commercial foods, because time can be measured independently. Emphasized instead is the lack of independent handling and preparation, assuming that such manipulation (and the greater variation and possibility of failure that results) is important for the development of real cooking skills and family-based food memories. Such foods are often highly processed, because increased processing tends to remove the possibility of independent decisions in dish completion. In Los Angeles, commercial foods were incorporated into weeknight dinner preparation in a variety of ways. Cooks may merely open up packages and follow the printed directions (one extreme) or prepare all dishes exclusively from raw ingredients (another extreme).

The aspect of time saving is emphasized by many studies on convenience food use (Verlegh and Candel, 1999). Florkowski et al. (2000, p. 173) argue that the “range of foods chosen from grocery shelves would be skewed toward processed or convenience foods if a household chooses to allocate only a limited amount of time for meal preparation”. Although very short preparation times were associated with heavy use of commercial foods in Los Angeles, increasing the time spent on meal preparation did not necessarily reduce or eliminate the use of commercial foods. These foods are therefore not described here as “convenience foods”, because they do not consistently reduce time inputs.

The results of this study suggest that the introduction and growing dependence upon commercial foods in the US diet may be a one-way street, as tastes, habits, and skills change in response to new goods and technology. Commercial foods reduced hands-on preparation by ten to 12 minutes on average, and were almost always necessary for hands-on preparation times under 20 minutes. These foods were also prominent in meals requiring considerable preparation time, however, indicating the importance of factors other than time. Commercial foods may reduce time investment at stages other than meal preparation, such as shopping.

Methodology

This research is part of a broader study on middle-class, dual-earner families in Los Angeles County, conducted by the Center on the Everyday Lives of Families (CELF) at the University of California-Los Angeles (Graesch et al., 2006; Ochs et al., 2006, 2007). Families volunteered for participation in the study. The requirements for inclusion were that both parents work at least 30 hours a week, that they own their own home and pay a mortgage, and that they have at least one child between eight to ten years of age. As with other research involving human subjects in the US, CELF's study was reviewed to make sure it complied with federal regulations for the protection of human research subjects (in this case, by an institutional review board overseen by UCLA's Office for Protection of Research Subjects).

Participants self-reported information such as income, personal background, and education level. All 32 households in the CELF study have an annual income of at least $45,000, which was approximately the median household income ($45,958) for Los Angeles County in 2004 (US Census Bureau, 2006). Most of the adult parents were born outside Los Angeles County and have a college degree if not a graduate degree. Ethnicities of individual parents include white of non-Hispanic descent (68 percent), African descent (8 percent), East or Southeast Asian descent (8 percent), Hispanic descent (8 percent), and South Asian descent (6 percent).

Fieldwork took place from January 2002 to December 2004 and included filming in the house on two weekdays while the family was there (before and after work/school activities). Two videographers recorded each family at home during the course of a week. On weekdays, these periods started when children left school and parents left work, and continued throughout the evening until the children retired to bed for the night.

In the course of filming, CELF researchers captured dinner preparations and dinner consumption on two weeknights for each family. These events were recorded by at least one of the two cameras in the house. Cameras were not positioned in a standard place in the home, but moved as necessary to record family activities and interactions. Of course, this known and planned observation may have influenced participant behavior. The extent to which this occurred is unknown, although it may be reduced by the fact that meal preparation was not a stated research focus at the time of filming. Neither participants nor researchers knew that use of commercial food would later be scrutinized, for example.

Of these 64 weeknight dinners, 45 (70 percent) were completely “home-cooked”, meaning they did not include food purchased from a restaurant as either take-out or delivery (Ochs et al., 2007). As discussed in the following, “home-cooked” meals may include extensive use of commercial food. Most of the remaining meals consisted of take-out food (14 percent) combined take-out food with some food prepared at home (5 percent), or were eaten in a restaurant (5 percent). Families ate at a relative's house for 3 percent of meals, and the last 3 percent were indeterminate (Ochs et al., 2007).

The data set for this paper is the 45 completely home-cooked weeknight meals, prepared for everyone that was home for dinner that night. Although not every family in the CELF study had at least one home-cooked meal on a weeknight during our observation period, 29 of the 32 families contributed one or two dinners to the sample. Observations of meal preparation (all conducted by the author) included the following: each task the cook performed, when and for how long he or she performed it, and what tools or appliances were used. All ingredients and dishes were identified, and meal consumption was also observed to make sure all dishes were recorded. From these observations, the following variables were extracted:

hands-on meal preparation time in minutes;
total meal preparation time in minutes;
the number of cooks working simultaneously;
number of dishes served in the entire meal;
composition of each dish, including the ingredients and manner of preparation, and whether it was a commercial dish or a modified commercial dish; and
home-cooked meal type, based on dish composition.

These variables are described in more detail in the following.

Timing (hands-on and total)

Meal preparation timing began when the cook began assembling the necessary ingredients or tools. In the CELF data, this often coincided with a verbal acknowledgement that the cook was going to start preparing dinner. Once timing started, it continued until all dishes were completely prepared (with the exception of breaks in cooking; see the following). This overall time is the total meal preparation time. This can be distinguished from the hands-on meal preparation time, or period when the cook is physically engaged in meal preparation. The hands-on time includes such activities as finding cooking utensils, washing or cutting ingredients, unwrapping packages, stirring a pot, and checking food to see if it is fully cooked. The hands-on time is the total time minus all interruptions or waiting periods that lasted longer than five minutes at a time. Shorter interruptions may have slowed down dinner preparation, but they were not subtracted from the hands-on time here. Timing ended once all dishes to be served were fully cooked or assembled.

Timing is focused on food preparation. Related activities necessary to eat dinner, such as cutting and serving the individual dishes and setting the table, are not considered meal preparation activities here for timing purposes.

All timing excludes breaks in cooking, or periods when cooking completely ceased. These are not periods when food is cooking unattended or the cook's attention is briefly turned to something else, which are included in the total meal preparation time. These breaks in cooking are significant interruptions that may occur between discrete steps of meal preparation. Examples include leaving the house to take a child to sports practice, or waiting for someone else who will finish food preparation.

Number of cooks

Any person who participated in food preparation, for any length of time, was considered to be a cook. When more than one cook worked simultaneously, time spent in meal preparation was not measured separately for each cook. The hands-on time includes any time that at least one person was physically engaged in food preparation.

Number and composition of dishes

Each dish prepared for dinner or consumed during dinner was counted. A dish is an individual menu item (e.g. salad, bread, casserole), often made from multiple ingredients. Dishes prepared for dinner but accidentally not served were still included in the dish count. The dish count also included additional items that were not part of the original meal plan, but sometimes found their way on to the table.

Dishes were further classified by their ingredients. If dishes included commercial food as ingredients, they were coded as commercial or modified commercial dishes. A commercial dish is made entirely from commercial food, as defined earlier. It is either purchased in its finished form or is finished by the home cook entirely according to package directions. Examples include pre-prepared frozen dishes such as pizza or pot pies. They also include mixes that either include all ingredients or dictate exactly what should be added to complete the dish, such as Hamburger Helper® (a stew or casserole mix, manufactured by Betty Crocker®, to which ground beef is added) and cornbread mix.

A modified commercial dish has at least one commercial ingredient, but the dish is finished independently by the cook, in a way not dictated solely by package directions. Examples include fish tacos made with frozen fish sticks, a baked pasta dish including commercial tomato sauce and pepperoni, or “pigs-in-a-blanket” (hot dogs wrapped in biscuit dough and baked).

The remaining dishes were either leftovers, originally prepared at an earlier date, or are composed entirely of raw ingredients. The distinction between the “commercial” ingredients in modified dishes and “raw” ingredients is not an easy distinction to make, given food availability and purchasing habits in the USA. The following items were not considered to be commercial ingredients for the purposes of this analysis: spices, seasonings, or marinades; dairy products, including cheese, yogurt, and whipped cream; dried pasta; ready-made tortillas from the grocery (although taco shells would have been commercial items); and frozen edamame (soybeans) in their pods. Commercial items include prepared meats such as hot dogs and pepperoni as well as vegetarian versions of burgers, hot dogs, and meatballs. They also include purchased tomato or alfredo sauces, rice in flavored rice mixes and pre-measured boil-in bags (but not non-instant rice purchased loose in larger quantities), and frozen or canned vegetables (which are processed to the extent that they may be simply heated and served, without washing, trimming, or adding other ingredients).

Home-cooked meal type

Percentages of commercial dishes and modified dishes in individual meals were plotted (Figure 1) to look for clusters that could be considered different types of meals. Breaks in the distribution of data points were used to divide home-cooked meals into three categories, as illustrated in Figure 1: limited use of commercial food: limited or no use of commercial dishes (<20 percent commercial dishes, ≤ 50 percent modified dishes); some use of commercial food: some use of commercial dishes (20-50 percent commercial dishes; if 50 percent of dishes were commercial, 0 percent of dishes were modified), and extensive use of commercial food: extensive or complete reliance on commercial dishes (≥50 percent commercial dishes; if 50 percent of dishes were commercial, additional dishes were modified).

Sample size and statistical analyses

All 45 completely home-cooked meals from 29 families are used to describe meal composition in the following. Only 40 meals from 26 families were used in the statistical analyses because of timing problems in five meals.

Preparation time can be treated as a continuous variable, or may be grouped into “short” (1st quartile and lower) and “long” (3rd quartile and higher) times to see which variables are associated with extremes in preparation time. Both approaches are used in the following. Although actual times in minutes are provided in descriptive summaries, the log values of times were used in ANOVA and Fisher's PLSD tests to normalize the distributions. The level of statistical significance throughout this paper is 0.05. All times were rounded to the nearest minute for analysis.

Results

All 45 meals include 14 meals (31 percent) with limited use of commercial food, 20 meals (44 percent) with some use, and 11 meals (24 percent) with extensive use. The 45 meals contained 156 dishes, including 56 (36 percent) commercial and 18 (12 percent) modified dishes. Most meals (n=37; 82 percent), were prepared by only one cook at a time. Simultaneous food preparation by two people was observed in eight meals.

Five of these 45 meals were excluded from statistical analyses of time spent on meal preparation: four because filming began after the start of meal preparation, preventing an accurate time record, and one because it included a dish cooked in a slow cooker or crock-pot and had an unusually long total preparation time. The remaining 40 meals contain 11 limited meals (28 percent), 18 some meals (45 percent), and 11 extensive meals (28 percent). The cooks prepared an average of three dishes per meal, spending an average of 34 minutes hands-on time and 52 minutes total time for the entire meal (Tables I and II).

The commercial dishes within these meals, of course, were prepared according to package directions. For other dishes, cooks only rarely referred to cookbooks, which often provide suggested menus in addition to individual recipes. This suggests that very few of the menus and dishes were taken directly from cookbooks. Only three cookbooks were observed during meal preparation, used by three different families to prepare three different meals.

Meal structure and components

Almost all observed meals contained at least one protein (98 percent of meals), one starch (98 percent), and one vegetable (84 percent) as a side dish or green salad or both. Salads with green leafy vegetables and desserts appeared in less than a quarter of the weeknight meals. The definitions of protein, starch, and vegetable used here are based on definitions from the United States Department of Agriculture (USDA, 2006) and discussed further in the following.

Almost all (98 percent) of the meals included a main dish with protein. This category combines foods in the USDA “meat and beans” and “milk” groups. In our sample of 45 home-cooked meals, the dominant protein was meat or poultry (80 percent), fish (7 percent), a commercial vegetarian meat substitute (4 percent), or cheese. Cheese, although it was an ingredient in many dishes, was the primary protein source in 7 percent of meals and offered as alternative main course to a vegetarian meat substitute in 4 percent. (The remaining meal was pasta with a vegetarian tomato sauce.)

Other possible protein sources (e.g. legumes, nuts, or eggs) were not used in an unprocessed form as the primary protein source in any of the observed home-cooked meals, although they may have appeared in commercial vegetarian meat substitutes. This was true even for the three vegetarian families in the study. For this reason, dry beans and peas were counted in the vegetable group, following the USDA recommendation for people who frequently eat meat, poultry, and fish.

Starch was as universal as protein, occurring in 98 percent of the meals. This category includes items in the USDA “grains” group. It also includes “starchy vegetables” as defined by the USDA, such as corn, green peas, and potatoes. Grains appeared as rice, yeast and quick breads, dumpling or pot sticker wrappers, tortillas, and pasta. The starchy vegetables in our sample are corn, green peas, and potatoes. Potatoes were usually not served with other starches, but corn or green peas were almost always served with other starches.

Vegetables include all foods in the USDA “vegetable” group, including dry beans and peas, excluding starchy vegetables. Although they are extremely common, vegetables seem more dispensable than other elements of the meal. They were served in 38 of 45 meals (84 percent) in one or more of the following forms: a side dish, usually cooked (16 meals), a salad with green leafy vegetables (13 meals), or ingredients in the main course (23 meals). The starchy vegetables green peas and corn are the only vegetables in three meals. They are excluded from the following discussion because they are discussed above as starches. If included, they would bring the number of meals with vegetables to 41 (91 percent).

Fruit was served with ten meals (22 percent) and included commercial applesauce as well as fresh fruit in the form of cantaloupe, grapes, pineapple, strawberries, and watermelon (fruits such as olives and avocados occasionally appeared in or with salads and are not included here). Other dessert foods were only served in four meals (9 percent), including in one meal with fruit. These foods were usually baked goods and were almost never made at home, or were baked using commercial packaged dough.

As noted previously (see Table II), the number of dishes ranges from one to seven. Only one one-dish meal (tacos) was served. Two factors help explain the meals with large numbers of dishes. One is the inclusion of leftovers as one or more extra dishes, which occurred in 11 meals (24 percent) of all 45 home-cooked meals. Another is the attempt to serve everyone at least one dish that he or she will eat. Extra dishes, whether leftovers or not, were added to 15 meals (33 percent) explicitly to accommodate the tastes of a particular family member.

Distributions of the numbers of dishes (Figure 2) suggest an upper limit of three dishes when commercial foods are limited or absent. Only 18 percent of the 11 “limited” meals contain more than three dishes, compared to 52 percent of the other meals, which incorporate commercial food. This difference is significant (p=<0.05) according to a G-squared test (df =1, G-squared=3.949, p=0.05). The two “limited” meals with more than three dishes include one meal with seven dishes, most added at the last minute by family members other than the cook. The other meal has four dishes; the cook, who grew up in India and still prepares Indian foods, relies heavily on preparing food ahead and using leftovers to cut preparation time.

Preparation time

The 40 meals in the statistical analysis contained an average of 3.6 dishes and required an average of 34 minutes hands-on time and 52 minutes total time to prepare (see Table I). The averages obscure considerable variation (Figure 3), some of it related to the use of commercial food. Preparation time (both hands-on and total time) does not vary significantly for meals with limited or no commercial food and meals with some use of commercial food. Extensive use of commercial food reduces hands-on time but not total meal preparation time. Cooks relying on commercial foods do not consistently get dinner on the table more quickly, if one focuses on total time.

About half of all preparation times can be described as “short” (equal to or less than the 1st quartile) or “long” (equal to or greater than the 3rd quartile). For hands-on time, short times are 13 to 22 minutes, and long times are 41 to 72 minutes. For total time, short times are 20 to 32 minutes, and long times are 68 to 127 minutes (Figure 4).

Meals with more dishes did not take more time to prepare in this study. When meals were divided into three groups based on the number of dishes (one to two, three to four, and five+ dishes), neither hands-on (p=0.88) nor total (p=0.081) time varied significantly (p=<0.05) by the number of dishes, according to an ANOVA test. As previously noted, meals with greater numbers of dishes often include dishes that are prepared extremely quickly from commercial foods.

Extensive reliance on commercial food reduces hands-on preparation time, bringing it to an average of 26 minutes (see Table I). This is significantly (p=<0.05) lower than average times with limited (p=0.05) or some (p=0.04) use of commercial food, according to a Fisher's PLSD test. The time saved is approximately ten to 12 minutes. Although the mean total time is also lower with extensive reliance on commercial time, the differences with limited (p=0.23) or some (p=0.12) use of commercial food are not statistically significant (p=<0.05), according to a Fisher's PLSD test.

Extensive use of commercial food is almost always necessary for very short hands-on preparation times. The group with short hands-on times (13-22 minutes) is dominated by meals with extensive use of commercial food, and only includes one meal with limited or no commercial food (Table III). This meal, with a hands-on time of 16 minutes, consisted of leftover turkey soup and toast. Differences in meal composition between short and long hands-on times are statistically significant (p=<0.05), according to a G-squared test (df =2, G-squared=6.395, p=0.004). This pattern is not related to the number of dishes served; the mean number of dishes does not vary significantly (p=<0.05) between short hands-on times (3.5 dishes) and long hands-on times (3.2 dishes), according to a Fisher's PLSD test (p=0.61).

To reduce hands-on time to the absolute minimum, cooks usually need to rely heavily on commercial foods. If total preparation time is the focus, then either commercial or non-commercial foods may be used to prepare a meal quickly. No significant differences (p=<0.05) in reliance on commercial food were found between meals with short and long total preparation times (df =2, G-squared=2.348, p=0.31).

Discussion

The ideal or stereotypical meal in the USA today contains at least three dishes, one each of protein, starch, and vegetables. This is similar to the A+2B formula, with one protein main course and two side dishes, described for British meals (Douglas, 1972; Douglas and Nicod, 1974). Cookbooks used by families in the study (e.g. Mills and Ross, 1997, p. 306) as well as other cookbooks that assume little if any prior knowledge of cooking (e.g. “The Complete Idiot's Guide to 20-Minute Meals”; Dimmick, 2003, p. 133) promote this notion of a balanced meal. In practice, the three components of a balanced meal can be combined so that one dish, for example, provides both protein and vegetables and the starch is served as a second dish. This is a similar, if simpler, meal structure than that proposed in the “Better Homes and Gardens” cookbooks from 1930 and 1953. Those suggested dinners contained a protein (which was always meat), a starch, and a vegetable side dish as well as a separate salad and a dessert (Levenstein, 1993, p. 119). Weeknight dinners in the CELF sample largely conformed to this pattern, at least in meal content if not number of dishes.

Families in our study often served extra dishes to please individual family members who did not like the rest of the meal. In their study of Italian-American families, Goode et al. (1984, p. 199) noted, “The way of handling the strong likes and dislikes of non-influential household members is to make them supplementary dishes or serve them leftovers” and argue that it is “a distinctly modern American pattern”. Other researchers have found that such accommodations to individual tastes are an accepted part of meal preparation in Italy, although they are often considered frustrating or annoying in the USA (Ochs et al., 1996). Some parents in this study approached such requests with flat refusal or open resentment while acquiescing.

In the CELF sample, meals dominated by commercial foods required on average ten to 12 minutes less hands-on time to prepare than meals with only some or limited use of commercial foods. No significant difference in total meal preparation time was found between meals prepared with and without commercial foods. Even minimal hands-on time savings may justify the extra expense of commercial food for some cooks. “When we're exhausted”, write Mills and Ross (1997, p. xv), “if the 20 minutes we bargained to spend cooking starts turning into 35, that's enough to finish us off”. The cooks in this study almost always used commercial food to reduce hands-on time to 20 minutes or less. On the other hand, many families in this study were willing to spend much more time than this cooking weeknight dinners.

Previous research on commercial food, usually labeled “convenience food”, addresses its use in relation to time expenditures (Verlegh and Candel, 1999; Florkowski et al., 2000). Economists treat time as a limited resource, like money, and have argued that the time spent in meal preparation is “determined by the opportunity cost of time, preference between market and nonmarket goods and leisure, and household production technology” (Florkowski et al., 2000, p. 173). In practice, “preference” may encompass other possible returns besides money, material goods, or time for leisure. Cooking may produce better-tasting food for less money, or may be oriented towards intangible rewards such as prestige, affection, approval, and pride in fulfilling one's social or spiritual role (e.g. Counihan, 1988; Sered, 1988).

Individuals who share the economists' orientation may be more likely to weigh variables and make food-preparation decisions in the manner they describe. Some do not, even in well-developed market economies. Davies and Madran (1997) instead emphasize perceived time pressure, or perceived opportunity cost of time. Although perceived time pressure is obviously related to actual time pressure, it increases for those who see time primarily as chronological and quantifiable, rather than successional. In the successional perspective, the event is more important than the time it takes. Based on surveys from 221 households in the UK, Davies and Madran (1997, p. 85) conclude that those with a past (tradition-based) orientation are more likely to enjoy cooking and to see meals as important events. They suggest “some people (about one-third of households) have not, at least as yet, allowed the general time pressures of modern living to affect their attitudes to food” (Davies and Madran, 1997, p. 86).

CELF study participants were not systematically asked whether they enjoyed food preparation or if cooking and eating at home were important to them. The research cited above suggests that people who enjoy these activities or believe they are important will spend more time on them, perceiving less time pressure. When families in this study were willing to spend time on meal preparation, however, they still often included commercial foods in the meal.

In total, 44 percent of the home-cooked meals incorporated commercial dishes but were not dominated by them. These meals were closer to meals with limited commercial food in preparation times, but closer to meals with extensive commercial food in the number of dishes. Various cookbooks promote some use of commercial food as a time saving strategy (e.g. Mills and Ross, 1997, pp. 358-360; Dimmick, 2003, p. 57). This may be true in particular recipes, but in the CELF sample, dinners with some commercial food do not require significantly less time to prepare than dinners with limited or no commercial food. The meals prepared with some commercial food are different, however; dinners are more elaborate with more processed food and more dishes. The commercial ingredients often permit a more complicated dish or meal than would have been possible in the same amount of time using only raw ingredients. Their use within another recipe may remove a perceived stigma associated with convenience foods (Gofton, 1995; Warde, 1999) by eliminating time savings and transforming the commercial item into something “home-made.”

As meals consumed in restaurants become less common after a period of great frequency (Sloan, 2006), home cooks may be trying to satisfy family tastes acquired in restaurants, including the ability to accommodate individual tastes and provide a variety of side dishes. Certain tastes and types of foods served in restaurants are difficult to prepare at home without purchasing key commercial ingredients. This was openly acknowledged in our sample by one cook, who moved to the USA as an adult; she sought to make her home-made spaghetti sauce taste like the restaurant version by adding a little commercial sauce as a seasoning.

Commercial foods also permit a different approach to provisioning and kitchen management. Three potential changes are listed in the following list:

Commercial foods increase flexibility in time scheduling. They are durable, unlikely to go bad before the family decides to eat them. Large volumes of food can be purchased at less frequent intervals. The freezer, as Warde (1999, pp. 522-523) notes, is one of the machines that “neither save labour nor compress time but rather allow the reordering of sequential use of time … it too may reduce the frequency of shopping trips”. Many families in this study stockpiled food within their homes and purchased more food even when food storage areas were relatively full. A total of 25 percent of families in our sample kept a second refrigerator or freezer in the garage for extra food storage.
Commercial foods reduce the need for planning ahead before shopping. They may limit or eliminate the need for a grocery list, because dishes can be purchased without a known recipe or any ingredients list; any ingredients not included are listed on the package exterior.
Money can be substituted for practical knowledge. Increased dietary variety is possible without learning new cooking skills or thumbing through cookbooks; the packaged version contains all or almost all of the ingredients, some or most of the preparation steps are already performed, and any remaining tasks are clearly outlined for the home cook. In our study, commercial versions of new foods were favored over assembly at home using directions. Cookbooks were infrequently used during meal preparation, and handwritten recipe cards or recipes clipped from printed media were never observed.

Evidence suggests that people with more money buy more convenience foods, although the connection between less time and convenience foods is debated (Warde, 1999; Gronau and Hamermesh, 2003). Unfortunately these heavily processed foods are often higher in sugar, fat, or salt and lower in nutrients and fiber than the raw ingredients used to make them, potentially leading to obesity or other health problems for those that rely heavily upon them (Nestle, 2002).

Conclusions

Observers predicted that during the 1990s, cooking in the USA would become the “assembly of ready-to-eat offerings into a meal” (Edidin, 1989). Visser and Stark (1989) argue that the microwave is “transforming the way we eat”, and suggest that “by the year 2000, the great majority of adults will have become so dependent upon prepared microwaveable meals that they will no longer know how to cook a meal for themselves.”

The validity of these predictions depends upon one's definition of cooking. In 53 percent of the observed weeknight meals in Los Angeles, at least one parent in a household demonstrated some independent cooking ability by preparing a meal at home that was not dominated by commercial food. Only 22 percent of the weeknight meals, however, were prepared at home with little or no commercial food. If middle-class parents in Los Angeles know how to cook a wide range of dishes from raw ingredients, they take only limited opportunities to exercise those skills, at least on weeknights. Other US research indicates at least some children are not acquiring these skills at home; young college women in one study did not know how to prepare many foods commonly consumed in the USA (Soliah et al., 2006).

Simple meals with limited numbers of dishes can be prepared quickly. This study illustrates, through observation of real families in real-life situations, that raw ingredients need not substantially increase preparation times. The menus and shopping strategies are likely to change with an increased emphasis on raw ingredients. Not all families (or individuals within families) want to prepare or eat such meals; they prefer the more complex menus that are possible, with the same time expenditure, using commercial foods. Others may find that fresh raw ingredients require more shopping trips or more planning than they can invest, or they may lack the cooking skills to prepare them. As habits, food preferences, and cooking skills in the USA evolve over time, for some of us commercial foods may be firmly embedded in our diets, regardless of how long we decide to spend making dinner.

Figure 1Home-cooked meal type for 45 meals, defined by percentages of commercial and modified dishes in each meal

Figure 2Box-and-whisker plot showing the distribution of number of dishes for meals with limited (n=11), some (n=18), and extensive (n=11) reliance on commercial food

Figure 3Box-and-whisker plot showing the distribution of preparation times for meals with limited (n=11), some (n=18), and extensive (n=11) reliance on commercial food

Figure 4Box-and-whisker plot showing the distribution of hands-on and total preparation times for 40 meals

Table IHands-on and total preparation time for 40 home-cooked meals

Table IINumber of dishes for 40 home-cooked meals

Table IIIReliance on commercial food in meals with short and long hands-on preparation times

References

Calta, M. (2005), Barbarians at the Plate: Taming and Feeding the Modern American Family, Perigee Books, New York, NY, .