- Email: [email protected]
Proximity of snacks to beverages increases food consumption in the workplace: A field study
Appetite 103 (2016) 244e248
Contents lists available at ScienceDirect
Appetite journal homepage: www.elsevier.com/locate/appet
Proximity of snacks to beverages increases food consumption in the workplace: A field study € Chance b, Nathan Novemsky b, Ravi Dhar b, Ernest Baskin a, *, 1, Margarita Gorlin b, 1, Zoe Kim Huskey c, Michelle Hatzis c a b c
Department of Food Marketing, Erivan K. Haub School of Business, Saint Joseph's University, USA Department of Marketing, Yale School of Management, Yale University, USA Google, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 October 2015 Received in revised form 20 April 2016 Accepted 21 April 2016 Available online 23 April 2016
In an effort to bolster employee satisfaction, many employers provide free snacks at the office. Unfortunately, keeping employees happy can conflict with the goal of keeping them healthy, since increased snacking at work can contribute to overeating and obesity. Building on the growing body of research in choice architecture, we tested one factor that might influence snack consumption without impacting satisfaction: the relative distance between snacks and beverages. In a large field study at Google, we measured snack consumption when snacks were closer to or farther from beverages. We found that employees who used the beverage station closer to the snack station were more likely to take a snacke the likelihood of snacking increased from 12% to 23% for men and from 13% to 17% for women when the beverage station closest to the snack station was used. These results imply that employers and even families could reduce snack consumption easily, cheaply, and without backlash, by increasing the relative distance between beverages and snacks. © 2016 Elsevier Ltd. All rights reserved.
keywords: Nutrition Proximity Consumption Field experiment Obesity Choice architecture Snacking
1. Introduction Access to free snacks and beverages at work is an important factor in employee satisfaction. In a survey of over 1000 employees, those who were provided free food at work were 20% more likely to report being extremely or very happy with their jobs, compared to those working without free snacks. Another survey found that 60% of employees reported that having more food at the office would make them feel more valued and appreciated by their employer (Taber, 2014). It is no surprise, therefore, that offering free food at work is a growing trend among employers (Malcolm, 2015). Unfortunately, the proliferation of free food increases consumption and may be taking a toll on public health. Widespread obesity in the United States and elsewhere has become a critical and costly public health problem (Hedley et al., 2004), and snacking is an important contributing factor. Snacking
* Corresponding author. Department of Food Marketing, Erivan K. Haub School of Business, Saint Joseph's University, 5600 City Avenue, Philadelphia, PA 19131, USA. E-mail address: [email protected] (E. Baskin). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.appet.2016.04.025 0195-6663/© 2016 Elsevier Ltd. All rights reserved.
has become a “fourth meal,” accounting for approximately a quarter of daily calories (Nielsen, Siega-Riz, & Popkin, 2002), and employees report that having free food available at work makes it difficult to keep their weight down (Taber, 2014). Thus, employers find themselves in a double-bind: offer free food that makes employees happy but unhealthy, or eliminate the food perks and risk employee backlash. Trying to simultaneously avoid both of these traps means offering snacks while somehow dissuading employees from eating too much of them. The question is, how? Most workplace and public health efforts to encourage healthy eating rely on informational campaigns. However, information alone rarely changes how people eat. For example, although people believe calorie labeling influences what and how much they eat (Elbel, Kersh, Brescoll, & Dixon, 2009), most studies have found it does not (Liu, Roberto, Liu, & Brownell, 2012; Long, Tobias, Cradock, Batchelder, & Gortmaker, 2015; Swartz, Braxton, & Viera, 2011). Also, when the National Institute of Health spent hundreds of millions of dollars on the “5 a Day” campaign to increase fruit and vegetable consumption in the United States, it yielded no difference in the number of Americans eating five servings of fruits and vegetables a day, and overall fruit and vegetable consumption declined
E. Baskin et al. / Appetite 103 (2016) 244e248
slightly during the campaign (Casagrande, Wang, Anderson, & Gary, 2007). A similar “5 a Day” campaign in the United Kingdom produced similarly dismal results (Hope, 2010). Fortunately, there are other promising and simple ways to encourage healthy choices: by changing the context in which choices are made. Since environmental cues can strongly impact food consumption (Wansink, 2014), employers might influence their employees' health by altering the design and layout of their snack areas. We conducted a large-scale field experiment investigating one way employers might subtly discourage snacking by modifying the environment. Having observed in an office break room that most employees entering the break room took a beverage, and being aware of previous research showing that accessibility of tempting foods increases their consumption (Cohen & Farley, 2008; Davis & Carpenter, 2009; Meyers & Stunkard, 1980; Painter, Wansink, & Hieggelke, 2002; Wansink, Painter, & Lee, 2006), we surmised that increasing the distance between beverages and snacks might reduce snack consumption, gently nudging employees toward healthier choices. While prior research (Wansink, Painter, et al., 2006) has found the constant temptation of a candy jar on a secretary's desk led her to consume more candy than she did when the jar was placed across the room, constant temptations like the desktop candy jar are rare. More typically, temptations are brief and avoidable, not sitting within view and within arms reach, all day. Our research focused on the situation in which employees enter a break room for a beverage and are briefly tempted by the snacks. While common sense suggests that having tasty snacks close at hand will increase temptation and therefore consumption, it is also quite possible that once employees have already walked 100 feet or more to the break room, an additional few feet of relative distance may have no effect. Moreover, some surprising laboratory experiments have found that sometimes exposure to temptation can ironically improve selfcontrol. Exposure to snacks has been found to make existing health goals salient, prompting people to override their temptations in a process called counteractive self-control (Fishbach, Friedman, & Kruglanski, 2003). This process can occur even before any temptation is sought (Trope & Fishbach, 2000). Since these surprising effects were reported in lab studies controlling for many other factors, we expected the real-world effects of counteractive self-control would likely be overwhelmed by the temptation of a proximal snack. We hypothesized that proximity to beverages would increase snack consumption. While there is no study examining how relative proximity of food or temptations might interact with gender, some research does suggest that women and men might respond differently. Women tend to be more concerned with their physical appearance and more dissatisfied with their body weight and shape (Tiggemann & Rothblum, 1997). They also feel more conflicted about food (Rolls, Fedoroff, & Guthrie, 1991), have a stronger belief in the link between diet and health, and worry more about whether their diet is healthy (Rozin, Bauer, & Catanese, 2003). For women, negative feelings about their appearance result in more restrained eating (Tiggemann, 1994). In addition, women have higher levels of self-discipline than men in certain domains (Duckworth & Seligman, 2006; Tiggemann, 1994). For these reasons, we expected women would be more conscious of snacking decisions and less likely to respond impulsively to nearby snacks. As a result, we hypothesized women's snacking behavior would be less influenced than men's by proximity due to greater deliberation and less impulsive snacking. We also expected snacking to increase throughout the day as employees' self-control resources became depleted. People find it increasingly difficult to resist temptation as decisions and other work tasks deplete their self-control reserves (Muraven &
245
Baumeister, 2000). Even judges, who are generally trusted to make impartial and fair decisions, were found to exert less selfcontrol in parole hearings the longer they had worked without a break (Danziger, Levav, & Avnaim-Pesso, 2011). We tested whether this depletion effect would interact with relative proximity to understand when relative proximity matters most. Our study contributes to the growing behavioral science literature on “choice architecture,” examining how contextual cues can guide healthy choices. Previous interventions have included moving bottled water to eye level (Thorndike, Riis, Sonnenberg, & Levy, 2014); putting healthy foods at the beginning of a buffet line (Wansink & Hanks, 2013); and switching to smaller bowls and utensils for serving ice cream (Wansink, Van Ittersum, & Painter, 2006). While no research has investigated the effect of the proximity of snacks to beverages, on a macro level, proximity to fast food restaurants has been found to correlate with higher obesity rates (Davis & Carpenter, 2009). Similarly, food salience in one's pantry at home has been shown to increase its consumption (Chandon & Wansink, 2002). To know whether relative proximity affects snacking in a typical work environment, a new study was needed. We further understanding of how contextual cues impact food consumption in several important ways. First, we test whether the relative distance between snacks and beverages will reduce snacking. Our field experiment allows us to test distances relevant in the real world: not only to workplace kitchens, but also cafeterias, buffets, retail settings, college dining halls, and to home kitchen and eating areas. Second, we examine the effect of brief temptations rather than constant temptations like the candy dish. Third, we investigate how gender and time of day interact with proximity of snacks to beverages. 2. Materials and methods We conducted a field study in the highest-traffic “microkitchen” in the Google New York office. A micro-kitchen is a break room containing hot and cold beverage stations and a variety of free snacks. The micro-kitchen in which we ran this study had two entrances, and two identical beverage stations with refrigerators and coffee machines, one near (60 500 ) and the other far (170600 ) from a snack bar displaying snacks such as M&Ms, chocolates, nuts, cookies, granola bars, packaged chips and pretzels, and fresh fruit (see Fig. 1 for a bird's-eye view of the micro-kitchen layout). Snack and beverage choices were recorded by research assistants unobtrusively working on laptops in the common areas next to the micro-kitchen. They recorded all beverage choices and recorded snack choice only if the employee's first action had been to take a beverage. 1170 beverage choices (67.6% by male employees) were observed over the course of seven full work days. For each of these observations, the following were recorded: relative proximity of the beverage station (near or far), whether a snack was taken, gender, and time of day. This micro-kitchen is frequented by approximately 400 employees and it was not feasible to track individuals, so some individuals likely appear more than once in our dataset. The design was quasi-random in that employees generally entered the micro-kitchen from whichever entrance (left or right) was closest to their office, or to the location from which they were coming. 3. Results and discussion Results were analyzed using a binary logistic regression to determine the effects of proximity, gender, and time of day on employees' tendency to take snacks with their beverages. Time of
246
E. Baskin et al. / Appetite 103 (2016) 244e248
Fig. 1. Layout of micro-kitchen.
day was calculated as the number of minutes past 9 a.m. The logistic regression model with relative proximity (near or far), gender, and time of day as independent variables and snack (yes or no) as the dependent variable explained substantial variance in snacking behavior, c2 (3) ¼ 17.19, p < 0.001. 3.1. Relative proximity The probability of snacking increased by more than half when an employee visited the near beverage machine rather than the far one, B ¼ 0.625, Wald ¼ 14.41, p < 0.001. Of the employees who used the beverage station near the snacks, 20.5% took a snack, whereas only 12.2% of those who used the far beverage station did. Thus, increased relative proximity of beverages resulted in a 69% increase in the probability of snacking. Given that the field study design did not permit random assignment to beverage stations, we considered the possibility that selection effects might impact the results: employees already intending to snack might strategically choose the “near” beverage station. Anecdotally, employees had told us they used whichever micro-kitchen entrance and beverage station was more convenient to their office, but we conducted a separate observation to test whether the alternative explanation was likely. Strategic selection of the “near” beverage station could not be ruled out if a sizeable number of employees were observed either 1) walking past the “far” beverage station to use the “near” one, or 2) walking past the micro-kitchen from the outside and then doubling back to use the entrance by the “near” beverage station (see Fig. 1). Of the 515 employees observed, only one walked past the “far” station to use the “near” one, and only one walked past the micro-kitchen and then doubled back, used the “near” beverage station, and took a snack. With only two out of 515 observations displaying potentially strategic beverage station choice, this explanation cannot account for the observed effect, suggesting that relative proximity is indeed responsible for differences in snacking behavior between employees using different beverage stations.
3.2. Gender To determine whether gender affected snack consumption, we ran a binary logistic regression with relative proximity, gender, and a relative proximity x gender interaction as independent variables and snacking incidence as the dependent variable. The interaction was not significant (B ¼ 0.480, Wald ¼ 1.858, p ¼ 0.173). However, when we conducted separate analyses for men and women, we did see different patterns. Our planned contrasts showed that, whereas women were not significantly more likely to snack when taking a beverage from the near (versus far) station (N ¼ 378; 17.1% versus 13.2%, B ¼ 0.3, Wald ¼ 1.069, p ¼ 0.301), men were more likely to snack when taking a beverage from the near station, rather than the far one (N ¼ 790; 22.5% versus 11.7%, B ¼ 0.780, Wald ¼ 15.352, p < 0.001). To put these results into perspective, for a man who drinks three beverages a day, weighs about 180 pounds and leads a mostly sedentary lifestyle at work, using the near coffee machine rather than the far one (e.g. because the near station happens to be closer to his desk) can lead him to consume an additional 81 snacks a year, which, conservatively assuming 150 calorie snacks, could result in a weight gain of about 2.5 pounds in one year (Body Weight Planner, 2016; Hall et al., 2011). Of course, some employees would likely experience much more than this average weight gain.
3.3. Time of day To determine the effect of time of day on consumption of snacks, we ran a binary logistic regression with relative proximity, time of day (number of minutes elapsed since 9 a.m.) and relative proximity time of day interaction as independent variables, with snacking incidence as the dependent variable. There was no significant interaction between time of day and relative proximity (B ¼ < 0.001, Wald ¼ 0.005, p ¼ 0.95). However, when looking at main effects, time of day emerged as a marginally significant predictor of snacking incidence (B ¼ 0.001, Wald ¼ 2.690, p ¼ 0.1).
E. Baskin et al. / Appetite 103 (2016) 244e248
Employees were more marginally more likely to take snacks as the day progressed. 4. Conclusions Our results contribute to the growing body of work showing how minor environmental differences can exert a powerful influence on behavior. These small changes can be used to further public health (Roberto, Pomeranz, & Fisher, 2014; Wansink & Chandon, 2014). In this case, a few additional feet between snacks and beverages caused a large enough decrease in snacking to potentially result in meaningful weight loss over time. Furthermore, since repeated behaviors can often cause people to form habits (Ouellette & Wood, 1998), consuming snacksdor not consuming snacksdwith beverages could become habitual, carrying over to other environments as well. There are some limitations of the present study that can be addressed in future research. First, since all snacks were free, we are unable to speculate about whether and how relative proximity erodes self-control when the tempting items must be purchased. Second, assignment to the near or far beverage station was only quasi-randomdemployees used the station closer to their desks. Although employees on both sides of the micro-kitchen work in similar job functions, it is possible (though not likely) that they differed in some relevant but unobservable way. Third, future studies should look at other populations. While we do not expect Google employees to show unusual effects of relative proximity, gender, or time of day on snacking, future studies can test the robustness of these findings. While our study focused on relative proximity, future research can also test many other contextual interventions for reducing impulsive snacking in the workplace. For example, blocking the line of sight between beverages and snacks could reduce temptation much as relative proximity does. Stocking free snacks in vending machines instead of on counters or shelves in order to increase the time and thought required to obtain them could dissuade consumption. Additionally, employers can make healthy snacks easier to reach and unhealthy snacks harder to reach. Future research could also explore other factors related to workplace snacking. First, research has shown that obesity may spread through social networks (Hill, Rand, Nowak, & Christakis, 2010; VanderWeele, 2011). Therefore, it is possible that observing others' snacking behavior might influence one's own. Second, studies have found that the presence of family and friends prompt greater food intake, and that cross-gender influences are complex (De Castro, 1994). It would be useful to examine whether proximity interacts with these social influences to affect snacking. Our results can generalize to other settings in which people might consume excess snacks because those snacks happen to be located close to another high-traffic location such as a beverage, silverware, or condiment station. In fact, any situation in which people are free to take as much food as they choose might be susceptible to the phenomenon described here. Therefore, both families and employers should consider these factors when making snack placement decisions. Although employers may be attempting to help employees by placing drinks and snacks conveniently in the same area, they may be unwittingly hurting employees by undermining their selfcontrol and increasing their consumption of tempting snacks. Thus, while some suggest employers can better serve employees by making it convenient to eat at the workplace (Wanjek, 2005), our research implies that sometimes less convenience might be more helpful. In particular, relative distance between snacks and beverages may decrease snack consumption and improve employees' health. We hope our findings spark insights and new questions
247
beyond workplace kitchens, helping to inform the optimal design of many eating spaces, such as cafeterias, dining halls, home kitchens, and any other settings in which tempting food challenges our self-control. References Body Weight Planner 2016. In. http://www.niddk.nih.gov/health-information/ health-topics/weight-control/body-weight-planner/Pages/bwp.aspx: National Institute of Diabetes and Digestive and Kidney Diseases. Casagrande, S. S., Wang, Y., Anderson, C., & Gary, T. L. (2007). Have Americans increased their fruit and vegetable intake?: the trends between 1988 and 2002. American Journal of Preventive Medicine, 32, 257e263. Chandon, P., & Wansink, B. (2002). When are stockpiled products consumed faster? a convenienceesalience framework of postpurchase consumption incidence and quantity. Journal of Marketing Research, 39, 321e335. Cohen, D., & Farley, T. A. (2008). Eating as an automatic behavior. Preventing Chronic Disease, 5. Danziger, S., Levav, J., & Avnaim-Pesso, L. (2011). Extraneous factors in judicial decisions. Proceedings of the National Academy of Sciences, 108, 6889e6892. Davis, B., & Carpenter, C. (2009). Proximity of fast-food restaurants to schools and adolescent obesity. American Journal of Public Health, 99, 505. De Castro, J. M. (1994). Family and friends produce greater social facilitation of food intake than other companions. Physiology & Behavior, 56, 445e455. Duckworth, A. L., & Seligman, M. E. (2006). Self-discipline gives girls the edge: gender in self-discipline, grades, and achievement test scores. Journal of Educational Psychology, 98, 198. Elbel, B., Kersh, R., Brescoll, V. L., & Dixon, L. B. (2009). Calorie labeling and food choices: a first look at the effects on low-income people in New York City. Health Affairs, 28, w1110ew1121. Fishbach, A., Friedman, R. S., & Kruglanski, A. W. (2003). Leading us not into temptation: momentary allurements elicit overriding goal activation. Journal of Personality and Social Psychology, 84, 296. Hall, K. D., Sacks, G., Chandramohan, D., Chow, C. C., Wang, Y. C., Gortmaker, S. L., et al. (2011). Quantification of the effect of energy imbalance on bodyweight. The Lancet, 378, 826e837. Hedley, A. A., Ogden, C. L., Johnson, C. L., Carroll, M. D., Curtin, L. R., & Flegal, K. M. (2004). Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. Journal of the American Medical Association, 291, 2847e2850. Hill, A. L., Rand, D. G., Nowak, M. A., & Christakis, N. A. (2010). Infectious disease modeling of social contagion in networks. PLoS Computational Biology, 6, e1000968. Hope, J. (2010). Millions spent on 5-a-day mantra but now we're eating even LESS vegetables. Daily Mail. dailymail.co.uk. Liu, P. J., Roberto, C. A., Liu, L. J., & Brownell, K. D. (2012). A test of different menu labeling presentations. Appetite, 59, 770e777. Long, M. W., Tobias, D. K., Cradock, A. L., Batchelder, H., & Gortmaker, S. L. (2015). Systematic review and meta-analysis of the impact of restaurant menu calorie labeling. American Journal of Public Health, 105, e11ee24. Malcolm, H. (2015). Study: The key to happiness at work is free snacks (Vol. 2015). USA Today (USA Today). Meyers, A. W., & Stunkard, A. J. (1980). Food accessibility and food choice: a test of Schachter's externality hypothesis. Archives of General Psychiatry, 37, 1133e1135. Muraven, M., & Baumeister, R. F. (2000). Self-regulation and depletion of limited resources: does self-control resemble a muscle? Psychological Bulletin, 126, 247. Nielsen, S. J., Siega-Riz, A. M., & Popkin, B. M. (2002). Trends in energy intake in US between 1977 and 1996: similar shifts seen across age groups. Obesity Research, 10, 370e378. Ouellette, J. A., & Wood, W. (1998). Habit and intention in everyday life: the multiple processes by which past behavior predicts future behavior. Psychological Bulletin, 124, 54. Painter, J. E., Wansink, B., & Hieggelke, J. B. (2002). How visibility and convenience influence candy consumption. Appetite, 38, 237e238. Roberto, C. A., Pomeranz, J. L., & Fisher, J. O. (2014). The need for public policies to promote healthier food consumption: a comment on Wansink and Chandon (2014). Journal of Consumer Psychology, 24, 438e445. Rolls, B. J., Fedoroff, I. C., & Guthrie, J. F. (1991). Gender differences in eating behavior and body weight regulation. Health Psychology, 10, 133. Rozin, P., Bauer, R., & Catanese, D. (2003). Food and life, pleasure and worry, among American college students: gender differences and regional similarities. Journal of Personality and Social Psychology, 85, 132. Swartz, J. J., Braxton, D., & Viera, A. J. (2011). Calorie menu labeling on quick-service restaurant menus: an updated systematic review of the literature. International Journal of Behavioral Nutrition and Physical Activity, 8, 135. Taber, L. (2014). Free food at work is on the rise d but not everyone's happy. NY POST. Thorndike, A. N., Riis, J., Sonnenberg, L. M., & Levy, D. E. (2014). Traffic-light labels and choice architecture: promoting healthy food choices. American Journal of Preventive Medicine, 46, 143e149. Tiggemann, M. (1994). Gender differences in the interrelationships between weight dissatisfaction, restraint, and self-esteem. Sex Roles, 30, 319e330. Tiggemann, M., & Rothblum, E. D. (1997). Gender differences in internal beliefs about weight and negative attitudes towards self and others. Psychology of
248
E. Baskin et al. / Appetite 103 (2016) 244e248
Women Quarterly, 21, 581e593. Trope, Y., & Fishbach, A. (2000). Counteractive self-control in overcoming temptation. Journal of Personality and Social Psychology, 79, 493. VanderWeele, T. J. (2011). Sensitivity analysis for contagion effects in social networks. Sociological Methods & Research, 40, 240e255. Wanjek, C. (2005). Food at work: workplace solutions for malnutrition, obesity and chronic diseases. International Labour Organization. Wansink, B. (2014). Slim by design. HarperCollins. Wansink, B., & Chandon, P. (2014). Slim by design: redirecting the accidental drivers
of mindless overeating. Journal of Consumer Psychology, 24, 413e431. Wansink, B., & Hanks, A. S. (2013). Slim by design: serving healthy foods first in buffet lines improves overall meal selection. PloS one, 8, e77055. Wansink, B., Painter, J. E., & Lee, Y.-K. (2006). The office candy dish: proximity's influence on estimated and actual consumption. International Journal of Obesity, 30, 871e875. Wansink, B., Van Ittersum, K., & Painter, J. E. (2006). Ice cream illusions: bowls, spoons, and self-served portion sizes. American Journal of Preventive Medicine, 31, 240e243.
Contents lists available at ScienceDirect
Appetite journal homepage: www.elsevier.com/locate/appet
Proximity of snacks to beverages increases food consumption in the workplace: A field study € Chance b, Nathan Novemsky b, Ravi Dhar b, Ernest Baskin a, *, 1, Margarita Gorlin b, 1, Zoe Kim Huskey c, Michelle Hatzis c a b c
Department of Food Marketing, Erivan K. Haub School of Business, Saint Joseph's University, USA Department of Marketing, Yale School of Management, Yale University, USA Google, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 October 2015 Received in revised form 20 April 2016 Accepted 21 April 2016 Available online 23 April 2016
In an effort to bolster employee satisfaction, many employers provide free snacks at the office. Unfortunately, keeping employees happy can conflict with the goal of keeping them healthy, since increased snacking at work can contribute to overeating and obesity. Building on the growing body of research in choice architecture, we tested one factor that might influence snack consumption without impacting satisfaction: the relative distance between snacks and beverages. In a large field study at Google, we measured snack consumption when snacks were closer to or farther from beverages. We found that employees who used the beverage station closer to the snack station were more likely to take a snacke the likelihood of snacking increased from 12% to 23% for men and from 13% to 17% for women when the beverage station closest to the snack station was used. These results imply that employers and even families could reduce snack consumption easily, cheaply, and without backlash, by increasing the relative distance between beverages and snacks. © 2016 Elsevier Ltd. All rights reserved.
keywords: Nutrition Proximity Consumption Field experiment Obesity Choice architecture Snacking
1. Introduction Access to free snacks and beverages at work is an important factor in employee satisfaction. In a survey of over 1000 employees, those who were provided free food at work were 20% more likely to report being extremely or very happy with their jobs, compared to those working without free snacks. Another survey found that 60% of employees reported that having more food at the office would make them feel more valued and appreciated by their employer (Taber, 2014). It is no surprise, therefore, that offering free food at work is a growing trend among employers (Malcolm, 2015). Unfortunately, the proliferation of free food increases consumption and may be taking a toll on public health. Widespread obesity in the United States and elsewhere has become a critical and costly public health problem (Hedley et al., 2004), and snacking is an important contributing factor. Snacking
* Corresponding author. Department of Food Marketing, Erivan K. Haub School of Business, Saint Joseph's University, 5600 City Avenue, Philadelphia, PA 19131, USA. E-mail address: [email protected] (E. Baskin). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.appet.2016.04.025 0195-6663/© 2016 Elsevier Ltd. All rights reserved.
has become a “fourth meal,” accounting for approximately a quarter of daily calories (Nielsen, Siega-Riz, & Popkin, 2002), and employees report that having free food available at work makes it difficult to keep their weight down (Taber, 2014). Thus, employers find themselves in a double-bind: offer free food that makes employees happy but unhealthy, or eliminate the food perks and risk employee backlash. Trying to simultaneously avoid both of these traps means offering snacks while somehow dissuading employees from eating too much of them. The question is, how? Most workplace and public health efforts to encourage healthy eating rely on informational campaigns. However, information alone rarely changes how people eat. For example, although people believe calorie labeling influences what and how much they eat (Elbel, Kersh, Brescoll, & Dixon, 2009), most studies have found it does not (Liu, Roberto, Liu, & Brownell, 2012; Long, Tobias, Cradock, Batchelder, & Gortmaker, 2015; Swartz, Braxton, & Viera, 2011). Also, when the National Institute of Health spent hundreds of millions of dollars on the “5 a Day” campaign to increase fruit and vegetable consumption in the United States, it yielded no difference in the number of Americans eating five servings of fruits and vegetables a day, and overall fruit and vegetable consumption declined
E. Baskin et al. / Appetite 103 (2016) 244e248
slightly during the campaign (Casagrande, Wang, Anderson, & Gary, 2007). A similar “5 a Day” campaign in the United Kingdom produced similarly dismal results (Hope, 2010). Fortunately, there are other promising and simple ways to encourage healthy choices: by changing the context in which choices are made. Since environmental cues can strongly impact food consumption (Wansink, 2014), employers might influence their employees' health by altering the design and layout of their snack areas. We conducted a large-scale field experiment investigating one way employers might subtly discourage snacking by modifying the environment. Having observed in an office break room that most employees entering the break room took a beverage, and being aware of previous research showing that accessibility of tempting foods increases their consumption (Cohen & Farley, 2008; Davis & Carpenter, 2009; Meyers & Stunkard, 1980; Painter, Wansink, & Hieggelke, 2002; Wansink, Painter, & Lee, 2006), we surmised that increasing the distance between beverages and snacks might reduce snack consumption, gently nudging employees toward healthier choices. While prior research (Wansink, Painter, et al., 2006) has found the constant temptation of a candy jar on a secretary's desk led her to consume more candy than she did when the jar was placed across the room, constant temptations like the desktop candy jar are rare. More typically, temptations are brief and avoidable, not sitting within view and within arms reach, all day. Our research focused on the situation in which employees enter a break room for a beverage and are briefly tempted by the snacks. While common sense suggests that having tasty snacks close at hand will increase temptation and therefore consumption, it is also quite possible that once employees have already walked 100 feet or more to the break room, an additional few feet of relative distance may have no effect. Moreover, some surprising laboratory experiments have found that sometimes exposure to temptation can ironically improve selfcontrol. Exposure to snacks has been found to make existing health goals salient, prompting people to override their temptations in a process called counteractive self-control (Fishbach, Friedman, & Kruglanski, 2003). This process can occur even before any temptation is sought (Trope & Fishbach, 2000). Since these surprising effects were reported in lab studies controlling for many other factors, we expected the real-world effects of counteractive self-control would likely be overwhelmed by the temptation of a proximal snack. We hypothesized that proximity to beverages would increase snack consumption. While there is no study examining how relative proximity of food or temptations might interact with gender, some research does suggest that women and men might respond differently. Women tend to be more concerned with their physical appearance and more dissatisfied with their body weight and shape (Tiggemann & Rothblum, 1997). They also feel more conflicted about food (Rolls, Fedoroff, & Guthrie, 1991), have a stronger belief in the link between diet and health, and worry more about whether their diet is healthy (Rozin, Bauer, & Catanese, 2003). For women, negative feelings about their appearance result in more restrained eating (Tiggemann, 1994). In addition, women have higher levels of self-discipline than men in certain domains (Duckworth & Seligman, 2006; Tiggemann, 1994). For these reasons, we expected women would be more conscious of snacking decisions and less likely to respond impulsively to nearby snacks. As a result, we hypothesized women's snacking behavior would be less influenced than men's by proximity due to greater deliberation and less impulsive snacking. We also expected snacking to increase throughout the day as employees' self-control resources became depleted. People find it increasingly difficult to resist temptation as decisions and other work tasks deplete their self-control reserves (Muraven &
245
Baumeister, 2000). Even judges, who are generally trusted to make impartial and fair decisions, were found to exert less selfcontrol in parole hearings the longer they had worked without a break (Danziger, Levav, & Avnaim-Pesso, 2011). We tested whether this depletion effect would interact with relative proximity to understand when relative proximity matters most. Our study contributes to the growing behavioral science literature on “choice architecture,” examining how contextual cues can guide healthy choices. Previous interventions have included moving bottled water to eye level (Thorndike, Riis, Sonnenberg, & Levy, 2014); putting healthy foods at the beginning of a buffet line (Wansink & Hanks, 2013); and switching to smaller bowls and utensils for serving ice cream (Wansink, Van Ittersum, & Painter, 2006). While no research has investigated the effect of the proximity of snacks to beverages, on a macro level, proximity to fast food restaurants has been found to correlate with higher obesity rates (Davis & Carpenter, 2009). Similarly, food salience in one's pantry at home has been shown to increase its consumption (Chandon & Wansink, 2002). To know whether relative proximity affects snacking in a typical work environment, a new study was needed. We further understanding of how contextual cues impact food consumption in several important ways. First, we test whether the relative distance between snacks and beverages will reduce snacking. Our field experiment allows us to test distances relevant in the real world: not only to workplace kitchens, but also cafeterias, buffets, retail settings, college dining halls, and to home kitchen and eating areas. Second, we examine the effect of brief temptations rather than constant temptations like the candy dish. Third, we investigate how gender and time of day interact with proximity of snacks to beverages. 2. Materials and methods We conducted a field study in the highest-traffic “microkitchen” in the Google New York office. A micro-kitchen is a break room containing hot and cold beverage stations and a variety of free snacks. The micro-kitchen in which we ran this study had two entrances, and two identical beverage stations with refrigerators and coffee machines, one near (60 500 ) and the other far (170600 ) from a snack bar displaying snacks such as M&Ms, chocolates, nuts, cookies, granola bars, packaged chips and pretzels, and fresh fruit (see Fig. 1 for a bird's-eye view of the micro-kitchen layout). Snack and beverage choices were recorded by research assistants unobtrusively working on laptops in the common areas next to the micro-kitchen. They recorded all beverage choices and recorded snack choice only if the employee's first action had been to take a beverage. 1170 beverage choices (67.6% by male employees) were observed over the course of seven full work days. For each of these observations, the following were recorded: relative proximity of the beverage station (near or far), whether a snack was taken, gender, and time of day. This micro-kitchen is frequented by approximately 400 employees and it was not feasible to track individuals, so some individuals likely appear more than once in our dataset. The design was quasi-random in that employees generally entered the micro-kitchen from whichever entrance (left or right) was closest to their office, or to the location from which they were coming. 3. Results and discussion Results were analyzed using a binary logistic regression to determine the effects of proximity, gender, and time of day on employees' tendency to take snacks with their beverages. Time of
246
E. Baskin et al. / Appetite 103 (2016) 244e248
Fig. 1. Layout of micro-kitchen.
day was calculated as the number of minutes past 9 a.m. The logistic regression model with relative proximity (near or far), gender, and time of day as independent variables and snack (yes or no) as the dependent variable explained substantial variance in snacking behavior, c2 (3) ¼ 17.19, p < 0.001. 3.1. Relative proximity The probability of snacking increased by more than half when an employee visited the near beverage machine rather than the far one, B ¼ 0.625, Wald ¼ 14.41, p < 0.001. Of the employees who used the beverage station near the snacks, 20.5% took a snack, whereas only 12.2% of those who used the far beverage station did. Thus, increased relative proximity of beverages resulted in a 69% increase in the probability of snacking. Given that the field study design did not permit random assignment to beverage stations, we considered the possibility that selection effects might impact the results: employees already intending to snack might strategically choose the “near” beverage station. Anecdotally, employees had told us they used whichever micro-kitchen entrance and beverage station was more convenient to their office, but we conducted a separate observation to test whether the alternative explanation was likely. Strategic selection of the “near” beverage station could not be ruled out if a sizeable number of employees were observed either 1) walking past the “far” beverage station to use the “near” one, or 2) walking past the micro-kitchen from the outside and then doubling back to use the entrance by the “near” beverage station (see Fig. 1). Of the 515 employees observed, only one walked past the “far” station to use the “near” one, and only one walked past the micro-kitchen and then doubled back, used the “near” beverage station, and took a snack. With only two out of 515 observations displaying potentially strategic beverage station choice, this explanation cannot account for the observed effect, suggesting that relative proximity is indeed responsible for differences in snacking behavior between employees using different beverage stations.
3.2. Gender To determine whether gender affected snack consumption, we ran a binary logistic regression with relative proximity, gender, and a relative proximity x gender interaction as independent variables and snacking incidence as the dependent variable. The interaction was not significant (B ¼ 0.480, Wald ¼ 1.858, p ¼ 0.173). However, when we conducted separate analyses for men and women, we did see different patterns. Our planned contrasts showed that, whereas women were not significantly more likely to snack when taking a beverage from the near (versus far) station (N ¼ 378; 17.1% versus 13.2%, B ¼ 0.3, Wald ¼ 1.069, p ¼ 0.301), men were more likely to snack when taking a beverage from the near station, rather than the far one (N ¼ 790; 22.5% versus 11.7%, B ¼ 0.780, Wald ¼ 15.352, p < 0.001). To put these results into perspective, for a man who drinks three beverages a day, weighs about 180 pounds and leads a mostly sedentary lifestyle at work, using the near coffee machine rather than the far one (e.g. because the near station happens to be closer to his desk) can lead him to consume an additional 81 snacks a year, which, conservatively assuming 150 calorie snacks, could result in a weight gain of about 2.5 pounds in one year (Body Weight Planner, 2016; Hall et al., 2011). Of course, some employees would likely experience much more than this average weight gain.
3.3. Time of day To determine the effect of time of day on consumption of snacks, we ran a binary logistic regression with relative proximity, time of day (number of minutes elapsed since 9 a.m.) and relative proximity time of day interaction as independent variables, with snacking incidence as the dependent variable. There was no significant interaction between time of day and relative proximity (B ¼ < 0.001, Wald ¼ 0.005, p ¼ 0.95). However, when looking at main effects, time of day emerged as a marginally significant predictor of snacking incidence (B ¼ 0.001, Wald ¼ 2.690, p ¼ 0.1).
E. Baskin et al. / Appetite 103 (2016) 244e248
Employees were more marginally more likely to take snacks as the day progressed. 4. Conclusions Our results contribute to the growing body of work showing how minor environmental differences can exert a powerful influence on behavior. These small changes can be used to further public health (Roberto, Pomeranz, & Fisher, 2014; Wansink & Chandon, 2014). In this case, a few additional feet between snacks and beverages caused a large enough decrease in snacking to potentially result in meaningful weight loss over time. Furthermore, since repeated behaviors can often cause people to form habits (Ouellette & Wood, 1998), consuming snacksdor not consuming snacksdwith beverages could become habitual, carrying over to other environments as well. There are some limitations of the present study that can be addressed in future research. First, since all snacks were free, we are unable to speculate about whether and how relative proximity erodes self-control when the tempting items must be purchased. Second, assignment to the near or far beverage station was only quasi-randomdemployees used the station closer to their desks. Although employees on both sides of the micro-kitchen work in similar job functions, it is possible (though not likely) that they differed in some relevant but unobservable way. Third, future studies should look at other populations. While we do not expect Google employees to show unusual effects of relative proximity, gender, or time of day on snacking, future studies can test the robustness of these findings. While our study focused on relative proximity, future research can also test many other contextual interventions for reducing impulsive snacking in the workplace. For example, blocking the line of sight between beverages and snacks could reduce temptation much as relative proximity does. Stocking free snacks in vending machines instead of on counters or shelves in order to increase the time and thought required to obtain them could dissuade consumption. Additionally, employers can make healthy snacks easier to reach and unhealthy snacks harder to reach. Future research could also explore other factors related to workplace snacking. First, research has shown that obesity may spread through social networks (Hill, Rand, Nowak, & Christakis, 2010; VanderWeele, 2011). Therefore, it is possible that observing others' snacking behavior might influence one's own. Second, studies have found that the presence of family and friends prompt greater food intake, and that cross-gender influences are complex (De Castro, 1994). It would be useful to examine whether proximity interacts with these social influences to affect snacking. Our results can generalize to other settings in which people might consume excess snacks because those snacks happen to be located close to another high-traffic location such as a beverage, silverware, or condiment station. In fact, any situation in which people are free to take as much food as they choose might be susceptible to the phenomenon described here. Therefore, both families and employers should consider these factors when making snack placement decisions. Although employers may be attempting to help employees by placing drinks and snacks conveniently in the same area, they may be unwittingly hurting employees by undermining their selfcontrol and increasing their consumption of tempting snacks. Thus, while some suggest employers can better serve employees by making it convenient to eat at the workplace (Wanjek, 2005), our research implies that sometimes less convenience might be more helpful. In particular, relative distance between snacks and beverages may decrease snack consumption and improve employees' health. We hope our findings spark insights and new questions
247
beyond workplace kitchens, helping to inform the optimal design of many eating spaces, such as cafeterias, dining halls, home kitchens, and any other settings in which tempting food challenges our self-control. References Body Weight Planner 2016. In. http://www.niddk.nih.gov/health-information/ health-topics/weight-control/body-weight-planner/Pages/bwp.aspx: National Institute of Diabetes and Digestive and Kidney Diseases. Casagrande, S. S., Wang, Y., Anderson, C., & Gary, T. L. (2007). Have Americans increased their fruit and vegetable intake?: the trends between 1988 and 2002. American Journal of Preventive Medicine, 32, 257e263. Chandon, P., & Wansink, B. (2002). When are stockpiled products consumed faster? a convenienceesalience framework of postpurchase consumption incidence and quantity. Journal of Marketing Research, 39, 321e335. Cohen, D., & Farley, T. A. (2008). Eating as an automatic behavior. Preventing Chronic Disease, 5. Danziger, S., Levav, J., & Avnaim-Pesso, L. (2011). Extraneous factors in judicial decisions. Proceedings of the National Academy of Sciences, 108, 6889e6892. Davis, B., & Carpenter, C. (2009). Proximity of fast-food restaurants to schools and adolescent obesity. American Journal of Public Health, 99, 505. De Castro, J. M. (1994). Family and friends produce greater social facilitation of food intake than other companions. Physiology & Behavior, 56, 445e455. Duckworth, A. L., & Seligman, M. E. (2006). Self-discipline gives girls the edge: gender in self-discipline, grades, and achievement test scores. Journal of Educational Psychology, 98, 198. Elbel, B., Kersh, R., Brescoll, V. L., & Dixon, L. B. (2009). Calorie labeling and food choices: a first look at the effects on low-income people in New York City. Health Affairs, 28, w1110ew1121. Fishbach, A., Friedman, R. S., & Kruglanski, A. W. (2003). Leading us not into temptation: momentary allurements elicit overriding goal activation. Journal of Personality and Social Psychology, 84, 296. Hall, K. D., Sacks, G., Chandramohan, D., Chow, C. C., Wang, Y. C., Gortmaker, S. L., et al. (2011). Quantification of the effect of energy imbalance on bodyweight. The Lancet, 378, 826e837. Hedley, A. A., Ogden, C. L., Johnson, C. L., Carroll, M. D., Curtin, L. R., & Flegal, K. M. (2004). Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. Journal of the American Medical Association, 291, 2847e2850. Hill, A. L., Rand, D. G., Nowak, M. A., & Christakis, N. A. (2010). Infectious disease modeling of social contagion in networks. PLoS Computational Biology, 6, e1000968. Hope, J. (2010). Millions spent on 5-a-day mantra but now we're eating even LESS vegetables. Daily Mail. dailymail.co.uk. Liu, P. J., Roberto, C. A., Liu, L. J., & Brownell, K. D. (2012). A test of different menu labeling presentations. Appetite, 59, 770e777. Long, M. W., Tobias, D. K., Cradock, A. L., Batchelder, H., & Gortmaker, S. L. (2015). Systematic review and meta-analysis of the impact of restaurant menu calorie labeling. American Journal of Public Health, 105, e11ee24. Malcolm, H. (2015). Study: The key to happiness at work is free snacks (Vol. 2015). USA Today (USA Today). Meyers, A. W., & Stunkard, A. J. (1980). Food accessibility and food choice: a test of Schachter's externality hypothesis. Archives of General Psychiatry, 37, 1133e1135. Muraven, M., & Baumeister, R. F. (2000). Self-regulation and depletion of limited resources: does self-control resemble a muscle? Psychological Bulletin, 126, 247. Nielsen, S. J., Siega-Riz, A. M., & Popkin, B. M. (2002). Trends in energy intake in US between 1977 and 1996: similar shifts seen across age groups. Obesity Research, 10, 370e378. Ouellette, J. A., & Wood, W. (1998). Habit and intention in everyday life: the multiple processes by which past behavior predicts future behavior. Psychological Bulletin, 124, 54. Painter, J. E., Wansink, B., & Hieggelke, J. B. (2002). How visibility and convenience influence candy consumption. Appetite, 38, 237e238. Roberto, C. A., Pomeranz, J. L., & Fisher, J. O. (2014). The need for public policies to promote healthier food consumption: a comment on Wansink and Chandon (2014). Journal of Consumer Psychology, 24, 438e445. Rolls, B. J., Fedoroff, I. C., & Guthrie, J. F. (1991). Gender differences in eating behavior and body weight regulation. Health Psychology, 10, 133. Rozin, P., Bauer, R., & Catanese, D. (2003). Food and life, pleasure and worry, among American college students: gender differences and regional similarities. Journal of Personality and Social Psychology, 85, 132. Swartz, J. J., Braxton, D., & Viera, A. J. (2011). Calorie menu labeling on quick-service restaurant menus: an updated systematic review of the literature. International Journal of Behavioral Nutrition and Physical Activity, 8, 135. Taber, L. (2014). Free food at work is on the rise d but not everyone's happy. NY POST. Thorndike, A. N., Riis, J., Sonnenberg, L. M., & Levy, D. E. (2014). Traffic-light labels and choice architecture: promoting healthy food choices. American Journal of Preventive Medicine, 46, 143e149. Tiggemann, M. (1994). Gender differences in the interrelationships between weight dissatisfaction, restraint, and self-esteem. Sex Roles, 30, 319e330. Tiggemann, M., & Rothblum, E. D. (1997). Gender differences in internal beliefs about weight and negative attitudes towards self and others. Psychology of
248
E. Baskin et al. / Appetite 103 (2016) 244e248
Women Quarterly, 21, 581e593. Trope, Y., & Fishbach, A. (2000). Counteractive self-control in overcoming temptation. Journal of Personality and Social Psychology, 79, 493. VanderWeele, T. J. (2011). Sensitivity analysis for contagion effects in social networks. Sociological Methods & Research, 40, 240e255. Wanjek, C. (2005). Food at work: workplace solutions for malnutrition, obesity and chronic diseases. International Labour Organization. Wansink, B. (2014). Slim by design. HarperCollins. Wansink, B., & Chandon, P. (2014). Slim by design: redirecting the accidental drivers
of mindless overeating. Journal of Consumer Psychology, 24, 413e431. Wansink, B., & Hanks, A. S. (2013). Slim by design: serving healthy foods first in buffet lines improves overall meal selection. PloS one, 8, e77055. Wansink, B., Painter, J. E., & Lee, Y.-K. (2006). The office candy dish: proximity's influence on estimated and actual consumption. International Journal of Obesity, 30, 871e875. Wansink, B., Van Ittersum, K., & Painter, J. E. (2006). Ice cream illusions: bowls, spoons, and self-served portion sizes. American Journal of Preventive Medicine, 31, 240e243.