Calorie Labeling and Food Choices: A First Look at the Effects on Low-Income People in New York City

Abstract

We examined the influence of menu calorie labels on fast food choices in the wake of New York City's labeling mandate. Receipts and survey responses were collected from 1,156 adults at fast-food restaurants in low-income, minority New York communities. These were compared to a sample in Newark, New Jersey, a city that had not introduced menu labeling. We found that 27.7 percent who saw calorie labeling in New York said the information influenced their choices. However, we did not detect a change in calories purchased after the introduction of calorie labeling. We encourage more research on menu labeling and greater attention to evaluating and implementing other obesity-related policies.

Full text

Calorie Labeling And Food
Choices: A First Look At The
Effects On Low-Income People
In New York City
Calorie information on menus appears to increase awareness of
calorie content, but not necessarily the number of calories people
purchase.
by Brian Elbel, Rogan Kersh, Victoria L. Brescoll, and L. Beth Dixon
ABSTRACT: We examined the influence of menu calorie labels on fast food choices in the
wake of New York City's labeling mandate. Receipts and survey responses were collected
from 1,156 adults at fast-food restaurants in low-income, minority New York communities.
These were compared to a sample in Newark, New Jersey, a city that had not introduced
menu labeling. We found that 27.7 percent who saw calorie labeling in New York said the in-
formation influenced their choices. However, we did not detect a change in calories pur-
chased after the introduction of calorie labeling. We encourage more research on menu la-
beling and greater attention to evaluating and implementing other obesity-related policies.
[Health Aff (Millwood). 2009;28(6):w1110–21 (published online 6 October 2009; 10.1377/
hlthaff.28.6.w1110)]
Several years after the u.s. surgeon general’s public warning of
an “obesity epidemic,” public policy responses have been patchwork and
partial.1Although more than 100 bills have been introduced since 2002, no
major legislation to address the problem has passed the U.S. Congress to date.2
States and metropolitan areas vary widely in the degree and nature of their legisla-
tive and regulatory activity.3Experts in the science and politics of nutrition have
reached some consensus around feasible policy options that could have an impact
on rising obesity rates.4However, few of these options have been implemented on
a scale that would permit systematic evaluation.
w1110 6 October 2009
Obesity
DOI 10.1377/hlthaff.28.6.w1110 ©2009 Project HOPE–The People-to-People Health Foundation, Inc.
Brian Elbel (brian .elbel@nyumc.org) is an assistant professor in the Division of General Internal Medicine at the
New York University (NYU) School of Medicine and the NYU Wagner School of Public Service, both in New York
City. Rogan Kersh is an associate professor and associate dean of the Wagner School. Victoria Brescoll is an
assistant professor in the Yale School of Management. Beth Dixon is an associate professor in the NYU Steinhardt
School of Culture, Education, and Human Development.

nCalorie labeling. One recently emergent and rapidly expanding policy to ad-
dressobesityratesiscalorielabeling(alsoreferredtoasmenulabeling).NewYork
City became the first U.S. jurisdiction to implement this legislation, on 19 July 2008.
Although the proposed regulatory details differ across localities, the statutes typi-
cally require restaurants with a certain number of locations in a city or state (rang-
ing from at least five to twenty; the number in New York City is fifteen) to visibly
post the caloric content of all regular menu items. In general, fast-food outlets must
post calorie labels on their menu boards; sit-down establishments are required to
list calories on the printed menu. In some cases, additional nutritional information is
required. NYC restaurants must list calories for all regularly available menu options,
using a typeface and format similar to the price or name of the item.5
Nutrition advocates view labeling as an important public policy tool to influ-
ence obesity at a population level, largely because of the strong link between fast-
food consumption and obesity.6More than thirty U.S. cities and states, including
the nation’s most populous city (New York) and state (California), have intro-
duced legislation to mandate menu labeling; thirteen bills had become law as of
this writing. At the federal level, consensus around a labeling bill seems to have
emerged in the Senate. This bill, which at the time of this writing has been rolled
into the larger set of bills addressing health reform, is very similar to the NYC leg-
islation.7
nPrevious studies. Little scientific evidence exists evaluating the influence of
menu labeling on fast-food choices.8–10 OnestudybytheNYCDepartmentofHealth
and Mental Hygiene examined food purchases at Subway restaurants that volun-
tarily posted calorie information in advance of mandatory labeling.11 They found that
Subway customers who saw the information (32 percent of respondents) consumed
fifty-two fewer calories, on average. The study could not account for health-
conscious consumers who might have been more likely to notice calorie information
and therefore purchased fewer calories because of their underlying preferences. A
recent experiment using random assignment of consumers in a nonrestaurant set-
ting found that menu labeling did not decrease calories ordered or consumed, even
among those who reported noticing the calorie information. In fact, that study
found some evidence that males ordered more calories when labels were present.12 A
second experiment examining calorie labeling on a printed menu found that label-
ing was effective in altering food consumed, but only when coupled with informa-
tion indicating that 2,000 was the recommended daily allowance of calories.13 Fi-
nally, a few studies have examined menu labeling in a cafeteria setting14–17 or via
hypothetical-choice experiments.18, 19 These studies found inconsistent and gener-
ally weak results from menu labeling.
nOur study. Using data collected before and after labeling was introduced in
New York City and a comparison location (Newark, New Jersey), we examined the
influence of calorie labeling on food choices. Given the increased risk of obesity and
related health problems associated with low-income and racially/ethnically diverse
Food Choices
HEALTH AFFAIRS ~ Web Exclusive w1111

populations,wefocusedourattentiononthesegroups.
20, 21 In addition to analyzing
calories purchased at fast-food restaurants, we also examined the percentage of con-
sumers who reported noticing and responding to calorie information.
Given the severe nature of this public health problem, careful scientific evalua-
tions of policy solutions are incredibly important. There are many policy propos-
als ranging from educational interventions to attempts to change the built envi-
ronment to make physical activity the “default” behavior in cities and states.
However, almost none of these policy interventions has actually been imple-
mented.4Calorie-labeling policies are among the first obesity policies to be widely
embraced. Yet we have virtually no data outside of the laboratory to examine
whether these policies are effective and, in particular, whether they are effective
among the most vulnerable populations.22 The study reported in this paper is the
first to evaluate the effectiveness of this policy since its introduction.
Study Data And Methods
nChoice of cities. We chose New York City because it is the first site in the
country to have introduced calorie labeling. We selected Newark as the control city
because (1) it has not introduced calorie labeling; (2) its urban characteristics and
demographics are similar to those of New York City; and (3) it does not have a vast
number of daily commuters to New York City but is close enough to permit a rea-
sonably consistent comparison.
nChoice of neighborhoods and restaurants. We began by narrowing restau-
rants to those representing four of the largest fast-food chains located in New York
City and Newark: McDonald’s, Burger King, Wendy’s, and KFC. We targeted res-
taurants within lower-income demographic areas that largely consist of minority
groups, mostly African American and Latino. We used six sets of population-level
characteristics to match two restaurants from the same chain in NYC neighbor-
hoods with one restaurant of the same chain in the Newark city limits: population
size, age, race/ethnicity, poverty level, obesity rates, and diabetes rates. We also at-
tempted to match key structural or geographic characteristics in our restaurant
pairings (for example, location relative to public transportation; proximity to large
apartment complexes, hospitals, or other institutions; and location in a downtown
area). After minimal restaurant substitutions, we were left with five restaurants in
NewarkandfourteeninNewYork(fiveWendy’s,eightMcDonald’s,threeBurger
King, and three KFC). In New York City, our data collection locations included four
of the five boroughs: the Bronx (specifically, the South Bronx), Brooklyn (central
Brooklyn), Manhattan (Harlem and Washington Heights), and Queens (the
Rockaways).23
nData collection. All restaurants were visited during lunch (generally 12:30–
3:00 p.m.) or dinner hours (generally 4:30–7:00 p.m.) for approximately 2.5 hours by
a research team of three to four people. Restaurants were visited on a Tuesday,
Wednesday, or Thursday (thereby avoiding days most likely to consist of “special” or
w1112 6 October 2009
Obesity

“treat” meals) over a two-week period beginning 8 July 2008—before calorie label-
ing was implemented in New York City.
We used a methodology similar to a “street-intercept” survey.24 Every customer
possible was approached as he or she entered the restaurant during our desig-
nated survey periods. Customers were asked to bring their receipts back and to
answer a set of questions for compensation of $2. Subjects were not told why the
receipts were being collected. It is difficult to assess cooperation rates with street-
intercept surveys, and we did not directly collect participation data. However, an-
other NYC study using the same method tracked the total number of customers
entering a fast-food restaurant during data collection (regardless of whether cus-
tomers were approached to take the survey) and found that 55 percent answered a
survey.11 This was consistent with our data collection.
Approximately four weeks after labeling wasintroducedinNewYorkCity,data
were again collected from the same restaurants, headed by the same research staff,
using the same methodology, on the same days of the week and during the same
time periods. To the extent that restaurants differ from each other in ways we can-
not observe, these differences should be minimized by collecting data from the
same locations both before and after labeling. Here we report on the results for re-
spondents age eighteen and older. Because food choices that parents make for
their children and that adolescents make for themselves are especially complex,
we examined these groups in other work.25, 26 We also limited our analysis to the
food that adults purchased for their own consumption, given the difficulty in
allocating calories from food items consumed by multiple people.
nMeasures. Nutritional value of food purchased. To gather valid nutrition data, study
staff obtained receipts indicating food items purchased for each participant’s own
consumption. Food items purchased, along with any modifications or additions (for
example, added cheese, regular or diet soda), were confirmed by study staff with
oralreview.Wethenusedthenutritiondataprovidedoneachfast-foodestablish
-
ment’s corporate Web site to manually calculate for each item purchased and for the
order as a whole the following nutritional information: calories, saturated fat, so-
dium, and sugar. We chose these nutrients based on their associations with obesity,
chronic disease, and overall health. All menu items and respective nutrition infor-
mation were entered into a spreadsheet; all items were then verified by a second
group of research assistants.
Additional data collected. After the food purchase details were confirmed, a short
survey was conducted that included respondent’s age, sex, race (African Ameri-
can/black, Latino, other race/white), education (high school or less, some college
or an associate degree, a bachelor’s degree or above—these data were not collected
at baseline), and whether the food was consumed in the restaurant or taken “to
go.” We also asked respondents (1) whether they noticed any calorie information
posted in the establishment; (2) if so, whether the information influenced their
Food Choices
HEALTH AFFAIRS ~ Web Exclusive w1113

food choices; and (3) whether this calorie information caused them to purchase
more or fewer calories.
nStatistical analysis. First, we examined mean differences for all nutrition
variables across the pre- and postlabeling period for New York City and Newark.
We present those values adjusted for age, sex, race/ethnicity, and whether the food
was eaten in the restaurant or taken “to go.”23
Second, we also focused on the proportion of our sample who viewed calorie la-
bels and the proportion who indicated that the information influenced their
choice. We present the results separately for males versus females, respondents
younger and older than age thirty-five, and respondents who were black and those
who were Latino.
Finally, we examined the influence of noticing calorie information and whether
respondents were influenced by calorie labels for the post-labeling sample in New
YorkCity(examinedasasetofdummyvariablesandpresentedasregression-
adjusted results). The study was reviewed by the institutional review board at the
NYU School of Medicine. All analyses were done with SAS version 9.1. Standard
errors were clustered at the restaurant level.
Study Findings
After excluding twenty-one receipts for which specific food items could not be
confirmed, we analyzed data from 1,156 receipts collected from adults for food
they purchased for themselves. As per our design, 71 percent of our sample was
surveyed in New York City (47 percent of these before calorie labeling and the rest
after) with the remainder in Newark. Approximately 38 percent of our sample was
male, with a mean age of thirty-eight. Those identifying themselves as black made
up 65.7 percent of the sample; Latinos made up 19.9 percent; and the remaining
14.4 percent consisted of other races, including those identifying themselves as
mixed race or white. Almost half of our postlabeling sample had only a high school
diploma or less. Within cities, our sample stayed consistent, with the exception of
a statistically significant increase in the proportion of black respondents in New-
ark (increasing from 74 percent prelabeling to 81 percent postlabeling). Our New-
ark sample was also slightly more likely to be black and less likely to be Hispanic
than our NYC sample.23
nNotice of and response to calorie labels. At baseline, the percentage of peo-
ple who saw calorie information available on posters, pamphlets, or food wrappers
did not differ between New York City and Newark (Exhibit 1). However, after calo-
rie labeling was instituted in New York City, the percentage of respondents who re-
ported noticing calorie information increased sharply in New York City—to 54 per-
cent—but not in Newark.
New York City also saw an increase in the percentage of people who reported
using this information and deciding as a result to purchase fewer calories. Newark
saw no such increases. Put differently, 27.7 percent of our post-labeling NYC sam-
w1114 6 October 2009
Obesity

ple who saw the calorie labeling indicated that the information influenced their
choices. Of these, approximately 88 percent indicated that they purchased fewer
calories in response to labeling.23
nInfluence of labeling on the nutrient content of purchased food. People in
New York City purchased a mean number of 825 calories before menu labeling was
introduced and 846 calories after labeling was introduced (Exhibit 2). The number
of calories purchased in Newark before and after labeling also did not appreciably
change (823 calories before labeling and 826 calories after). Similar results were
Food Choices
HEALTH AFFAIRS ~ Web Exclusive w1115
EXHIBIT 1
Study Respondents Who Indicated Noticing And Responding To Calorie Labels In New
York City And Newark, New Jersey, Fast-Food Restaurants, 2008
SOURCE: Authors’ data.
NOTES: New York City was the study site; Newark was the comparison site. For all three questions, the NYC prelabeling period
was different from the NYC postlabeling period ( < 0.05), and the NYC postlabeling period was different from the Newarkp
postlabeling period ( < 0.05). No other dif ferences were statistically significant. A version of this exhibit showing 95 percentp
confidence intervals is available online at http://content.healthaffairs.org/cgi/content/full/hlthaff.28.6.1110/DC2.
40
30
20
10
0
Percent
NYC Newark NYC
Before labeling
After labeling
Newark
Noticed calorie labels
NYC Newark
50
Indicated that labels
influenced choice
Purchased fewer calories
EXHIBIT 2
Regression-Adjusted Nutrient Content For Food Purchases In New York City And
Newark, New Jersey, Before And After Calorie Labeling In Restaurants, 2008
New York City Newark
Before labeling After labeling Before labeling After labeling
Number of calories
Saturated fat (grams)
Sodium (milligrams)
Sugar (grams)
825
11.7
1,414
42
846
10.9
1,450
41
823
11.9
1,369
41
826
11.9
1,502
33
SOURCE: Authors’ data.
NOTES: There were no statistically significant differences. A version of this exhibit showing 95 percent confidence intervals is
available online at http://content.healthaffairs.org/cgi/content/full/hlthaff.28.6.w1110/DC2.

foundforsaturatedfat,sodium,andsugar, with no appreciable or significant differ-
ences before or after labeling was instituted.23
nCalories purchased by various population groups. Exhibit 3 presents only
the results for calories and whether these results differ by sex, race, or age. We found
no evidence that any of these groups differed in their responses to labeling, com-
pared to the sample as a whole. In each case, we saw neither a difference between
the NYC and Newark samples nor a difference before or after labeling.23
nPostlabeling findings. We analyzed the number of calories purchased by (1)
those who did not notice the posted calorie labels; (2) those who did notice the la-
bels but indicated that they were not inclined to purchase fewer calories as a result;
and (3) those who noticed the labels and indicated that as a result, they purchased
fewer calories (Exhibit 4).
Wefirstnotethattheserelationshipsarenotcausal,giventhatseeingthelabels
(or not) could be correlated with other factors that induce people to purchase
more or fewer calories. We found nonsignificant decreases in calories purchased
for groups who indicated that the labels mattered to them (blacks and people un-
der age thirty-five), while for other groups (older than age thirty-five) we found
nonsignificant increases.23
Discussion
In our study of consumers from low-income, minority communities, calorie la-
beling increased the percentage of consumers who reported seeing calorie labels,
and thereby the number of people who reported that the information influenced
w1116 6 October 2009
Obesity
EXHIBIT 3
Calories Purchased By Various Subgroups In New York City And Newark, New Jersey,
Fast-Food Restaurants, Before And After Calorie Labeling Began, 2008
S
OURCE: Authors’ data.
NOTES: New York City was the study site; Newark was the comparison site. Regression adjusted for age, race/ethnicity, and
whether or not food was purchased “to go.” A version of this exhibit showing 95 percent confidence intervals is available online
at http://content.healthaffairs.org/cgi/content/full/hlthaff.28.6.w1110/DC2.
800
600
400
200
0
Calories
NYC Newark NYC
Before labeling
After labeling
Newark
Male
NYC Newark
1,000
Female Black
NYC Newark NYC Newark NYC Newark
Latino Below age 35 Age 35 or older

their food choices. This meaningful change as a result of labeling could “set the
stage” for a larger influence of calorie labeling as time and public policy progress.
However, we did not find evidence in our sample that menu labeling influenced
the total number of calories purchased at the population level. About half of the
NYC respondents in our postlabeling sample reported noticing calorie informa-
tion, and only a quarter of these reported that the information influenced their
food choices. Even those who indicated that the calorie information influenced
their food choices did not actually purchase fewer calories according to our data
collection. We note again that our study sample consisted primarily of racial and
ethnic minorities residing in relatively low-income areas; other groups may re-
spond differently to labeling.
In an ideal world, calorie labeling on menus and menu boards would have an
immediate and direct impact on everyone’s food choices. However, as has been
seen in previous attempts to change the behavior of vulnerable populations (for
example, smoking cigarettes), greater attention to the root causes of behavior or
multifaceted interventions, or both, will be necessary if obesity is to be greatly re-
duced in the overall U.S. population.27–29 Future policy development must con-
sider this broader perspective.
nStrengths of our study. Our study had several advantages over the limited
prior research on calorie labeling. First, we studied labeling as it was rolled out in
the “real world,” as opposed to a hypothetical or laboratory setting. Second, we were
able to verify food—and therefore calories—purchased by examining respondents’
food receipts instead of obtaining retrospective reports. Third, we sampled the same
Food Choices
HEALTH AFFAIRS ~ Web Exclusive w1117
EXHIBIT 4
Number Of Calories Purchased In The New York City Sample After Calorie Labeling
Began, In Response To Calorie Labeling, 2008
S
OURCE: Authors’ data.
NOTES: Regression adjusted for age, race/ethnicity, and whether or not food was purchased “to go.” A version of this exhibit
showing 95 percent confidence intervals is available online at http://content.healthaffairs.org/cgi/content/full/hlthaff.28.6
.w1110/DC2.
800
600
400
200
0
Calories Did not notice
labeling
Noticed but did
not influence
Male
1,000
Female Black
Full
sample
Latino Below
age 35
Age 35
or older
Noticed and purchased
lower-calorie food

restaurants both before and after the introduction of labeling, thereby limiting the
effects of differences across restaurants. Fourth, the time period under study was
relatively narrow. Given the many factors that could influence people’s food
choices,30, 31 a short study period allowed us to better attribute any change in calories
purchased to the introduction of labeling (although we also note that our time pe-
riod was a potential problem, as described below). Fifth and finally, we included
data from not only New York City but also a comparison group—a critical study
design to control for possible trends in food choice unrelated to the calorie labeling.
nStudy limitations. Our study also had several limitations that point to the
need for future research—and that also may contribute to why we found low con-
sumer responsiveness to labeling.
First, although our short study period (approximately four weeks) was also a
strength of our design, the effect of labeling might have been different had we col-
lected our postlabeling data later. To the extent that repeated exposure is neces-
saryforbehaviorchange,ourshort-termstudydoesnotreflectthelonger-runim
-
pact of labeling. However, consumers in our sample reported frequenting fast-
food restaurants approximately five times per week, which indicates that they
likely had repeated experiences with calorie labels before our follow-up data col-
lection. It is not clear whether continued extensive exposure beyond a month
would have made consumers more or less likely to respond to labels.
The timing of our postlabeling data collection also meant that the exact format-
ting of some labels was in flux. Although all of the locations we studied posted
calorie labels, New York City levied fines on restaurants that were not in full com-
pliance with regulations requiring a specific typeface and placement of the calorie
labels.32 Labeling that was in full compliance with the regulation could have
altered our findings.
Second, menu labels might need to be coupled with greater education regarding
caloric content. Although education alone has not been successful in altering obe-
sity in the past, it is possible that an appropriately funded educational campaign
surrounding calorie labeling might improve the efficacy of calorie labeling.33 New
York City initiated an educational campaign (after our data collection) that in-
formedresidentsthat“2,000caloriesadayisallmostadultsshouldeat.”
34
Third, we were not able to observe whether some consumers were avoiding
outlets that posted calorie labels, because we sampled only customers who en-
tered a fast-food location with labeling. If consumers are avoiding restaurants
with labels, attention must be paid to where they are going instead—whether to
restaurants with less- or more-healthful food—and what they are consuming at
these locations. It is important to note that numerous restaurants and food service
establishments are not chains; as a result, only 10 percent of NYC restaurants are
subject to the labeling legislation.5
Fourth, it is possible that with a larger sample we might have observed a reduc-
tion in the number of calories purchased. Even a reduction of fifty calories (equiva-
w1118 6 October 2009
Obesity

lent to one Oreo cookie) per restaurant visit, sustained over time, could translate
to weight loss and potential health benefits for some people.
Fifth, future work must focus on whether labeling might be more effective at al-
tering the food choices of other subgroups (for example, those who eat fast food
more or less often or come from other demographic groups). Attention should be
paid to both structural reasons (for example, consumers not seeing or under-
standing calorie information) as well as reasons related to behavioral economics
and the psychology of food decision making.35–38
nNeed for additional interventions. Eating behavior is notoriously resistant to
change.39 A large body of research has shown that weight-loss interventions de-
signed to educate people about healthful food choices are generally ineffective. Thus,
simply displaying information about the caloric value of various food options may
fail to translate into attitudinal, motivational, or—most importantly—behavioral
changes in line with choosing healthier food options. Menu labels may need to be
coupled with additional policy approaches.
At the same time, our study does not necessarily imply that labeling is an inef-
fective policy. On the contrary, we found that some subset of consumers used the
information to eat more healthfully. Calorie labeling could result in changes that
do not rely primarily on alterations in consumers’ food choices. Menu labeling reg-
ulations may encourage chain restaurants to offer more nutritious or otherwise
improved menu offerings, which could be profoundly influential. Public health
experts have shown that creating “default” incentives to improve well-being is es-
sential to improving public health. By indirectly influencing restaurants to offer
more lower-calorie items, menu labeling regulations could help encourage such
default options for consumers.40, 41 That said, one study has found that simply add-
ing healthier options to a menu can counterintuitively increase the proportion of
consumers who purchase less-healthful menu items.42
Menu labeling is an important first attempt to alter food en-
vironments on a large scale and could ultimately prove both beneficial to
health and cost-effective. However, we simultaneously encourage fur-
ther research on menu labeling and much greater attention to implementing and
evaluating other obesity-related policies. Given the scale, scope, and difficulty in
combating the problem of obesity, greater attention must be given to the overall
range of policy options and to ways of making nascent policies, such as menu la-
beling, optimally effective.
Food Choices
HEALTH AFFAIRS ~ Web Exclusive w1119

This paper was presented at the AcademyHealth Annual Research Meeting, June 2009, in Chicago, Illinois. The
research was funded by the Robert Wood Johnson Foundation Healthy Eating Research Program, the Yale Rudd
Center for Food Policy and Obesity, and the New York University Wagner Dean’s Fund. The authors thank Joyce
Gyamfi, Kevin Lyu, Kristin Van Busum, Melinda Newe, Courtney Abrams, and their research assistants for their
great efforts on this project. They are also grateful for comments received during presentations at AcademyHealth,
American Society of Reproductive Medicine, Harvard School of Public Health, Cornell University, Johns Hopkins
Bloomberg School of Public Health, New York University School of Medicine, Columbia University Mailman
School of Public Health, and Mt. Sinai School of Medicine. Finally, the authors thank Beth Weitzman, Tod
Mijanovich, Kelly Brownell, and Marion Nestle for their assistance with this project.
NOTES
1. U.S. Department of Health and Human Services. The surgeon general’s call to action to prevent and de-
crease overweight and obesity. Washington (DC): DHHS; 2001.
2. Burnett D. Fast-food lawsuits and the cheeseburger bill: critiquing Congress’s response to the obesity epi-
demic. Va J Soc Pol’y and L. 2007;14(3):357–417.
3. National Conference of State Legislatures. Obesity statistics in the United States [Internet]. Washington
(DC): 2009 Jul [cited 2009 Sep 11]. Available from: http://www.ncsl.org/programs/health/obesity.htm
4. Brescoll VL, Kersh R, Brownell KD. Assessing the feasibility and impact of federal childhood obesity pre-
vention policies. Ann Amer Acad Polit Soc Sci. 2008;615(1):178–94.
5. New York City Department of Health and Mental Hygiene. Notice of Intention to Repeal and Reenact Sec.
81.50 of the New York City Health Code. New York (NY): Department of Health and Mental Hygiene;
2007.
6. Center for Science in the Public Interest. Anyone’s guess: the need for nutritional labeling at fast-food and
other chain restaurants. Washington (DC): CSPI; 2003.
7. Shields T. Restaurants, senators agree on nutrition labeling law [Internet]. Washington (DC): Bloomberg;
2009 Jun 11 [cited 2009 Sep 11]. Available from: http://www.bloomberg.com/apps/news?pid=20601103&
sid=aVDObtLXFZIY
8. Harnack LJ, French SA. Effect of point-of-purchase calorie labeling on restaurant and cafeteria food
choices: a review of the literature. Int J Behav Nutr Phys Act. 2008;5:51.
9. Berman M, Lavizzo-Mourey R. Obesity prevention in the information age: caloric information at the point
of purchase. JAMA. 2008;300(4):433–5.
10. Larson N, Story M (University of Minnesota, Minneapolis, MN). Menu labeling: does providing nutrition
information at the point of purchase affect consumer behavior? [Internet]. Princeton (NJ): Robert Wood
Johnson Foundation; 2009 Jun [cited 2009 Sep 11]. Available from: http://www.healthyeatingresearch.org/
images/stories/her_research_briefs/her_menu_labeling_brief_06_29_09_final.pdf. Healthy Eating Research
Synthesis.
11. Bassett MT, Dumanovsky T, Huang C, Silver LD, Young C, Nonas C, et al. Purchasing behavior and calorie
information at fast-food chains in New York City, 2007. Am J Public Health. 2008;98(8):1457–9.
12. Harnack LJ, French SA, Oakes JM, Story MT, Jeffery RW, Rydell SA. Effects of calorie labeling and value
size pricing on fast food meal choices: results from an experimental trial. Int J Behav Nutr Phys Act.
2008;5:63.
13. Roberto C, Larsen P, Agnew H, Baik J, Brownell K. Evaluating the affect of menu labeling on food choices
and intake. Am J Public Health (forthcoming).
14. Milich R, Anderson J, Mills M. Effects of visual presentation of caloric values on food buying by normal
and obese persons. Percept Mot Skills. 1976;42(1):155–62.
15. Cinciripini PM. Food choice and eating behavior among obese, lean, and normal individuals. Behav Modif.
1984;8(3):425–43.
16. Mayer J, Brown T, Heins J, Bishop D. A multi-component intervention for modifying food selections in a
worksite cafeteria. J Nutr Educ. 1987;19(6):277–80.
17. Burton S, Creyer EH, Kees J, Huggins K. Attacking the obesity epidemic: the potential health benefits of
providing nutrition information in restaurants. Am J Public Health. 2006;96(9):1669–75.
w1120 6 October 2009
Obesity

18. Balfour D, Moody R, Wise A, Brown K. Food choice in response to computer-generated nutrition infor-
mation provided about meal selection in workplace restaurants. J Human Nutr Dietetic. 1996;9(3):231–7.
19. Yamamoto JA, Yamamoto JB, Yamamoto BE, Yamamoto LG. Adolescent fast food and restaurant ordering
behavior with and without calorie and fat content menu information. J Adolesc Health. 2005;37(5):397–
402.
20. Chou SY, Grossman M, Saffer H. An economic analysis of adult obesity: results from the Behavioral Risk
Factor Surveillance System. J Health Econ. 2004;23(3):565–87.
21. HedleyAA,OgdenCL,JohnsonCL,CarrollMD,CurtinLR,FlegalKM.Prevalenceofoverweightandobe
-
sity among U.S. children, adolescents, and adults, 1999–2002. JAMA. 2004;291(23):2847–50.
22. Krukowski RA, Harvey-Berino J, Kolodinsky J, Narsana RT, Desisto TP. Consumers may not use or under-
stand calorie labeling in restaurants. J Am Diet Assoc. 2006;106(6):917–20.
23. For greater detail on geographic areas of data collection, study sample, and noticing and respond ing to cal-
orie labels, as well as full sets of regression results, see the online Appendix, available at http://content
.healthaffairs.org/cgi/content/full/hlthaff.28.6.w1110/DC2.
24. Bernard HR. Research methods in anthropology. Lanham (MD): Altamira Press; 2005. p. 161.
25. Bassett R, Chapman GE, Beagan BL. Autonomy and control: the co-construction of adolescent food
choice. Appetite. 2008;50(2–3):325–32.
26. Contento IR, Williams SS, Michela JL, Franklin AB. Understanding the food choice process of adolescents
in the context of family and friends. J Adolesc Health. 2006;38(5):575–82.
27. Nestle M, Cowell C. Health promotion for low-income minority groups: the challenge for nutrition educa-
tion. Health Educ Res. 1990;5(4):527–33.
28. Tilford S. Evidence-based health promotion. Health Educ Res. 2000;15(6):659–63.
29. Kerner JF, Dusenbury L, Mandelblatt J. Poverty and cultural diversity: challenges for health promotion
among the medically underserved. Annu Rev Public Health. 1993;14:355–77.
30. Downs JS, Loewenstein G, Wisdom J. The psychology of food consumption: strategies for promoting
healthier food choices. Am Econ Rev: Papers and Proceedings. 2009;99(2):1–10.
31. Sobal J, Bisogni CA, Devine C, Jastran M. A conceptual model of the food choice process. In: Shepherd R,
Raats M, editors. The psychology of food choice. 1st ed. Cambridge (MA): CABI; 2006.
32. Cordova EB. Health department issues mass citations to food chains; McDonald’s and Dunkin’ amass most
calorie citations. Crain’s New York; 2008 Oct 20.
33. Nestle M. Food politics: how the food industry influences nutrition and health. Berkley and Los Angeles
(CA): University of California Press; 2002.
34. New York City Department of Health and Mental Hygiene. 2,000 calories a day is all most adults should
eat [Internet]. New York (NY): Department of Health and Mental Hygiene; [cited 2009 Sep 11]. Available
from: http://www.nyc.gov/html/doh/downloads/pdf/calories/Calorie-Posters.pdf
35. DeJoy DM. The optimism bias and traffic accident risk perception. Accid Anal Prev. 1989;21(4):333–40.
36. Weinstein N. Unrealistic optimism about future lifeevents.JPersSocPsychol.1980;39(5):806.
37. Kan K. Cigarette smoking and self-control. J Health Econ. 2007;26(1):61–81.
38. Khwaja A, Silverman D, Sloan F. Time preference, time discounting, and smoking decisions. J Health Econ.
2007;26(5):927–49.
39. Mann T, Tomiyama AJ, Westling E, Lew AM, Samuels B, Chatman J. Medicare’s search for effective obesity
treatments: diets are not the answer. Am Psychol. 2007;62(3):220–33.
40. Just DR, Mancino L, Wansink B. Could behavioral economics help improve diet quality for nutrition assis-
tance program participants? [Internet]. Washington (DC): U.S. Department of Agriculture; 2007 [cited
2009 Sep 11]. Available from: http://www.ers.usda.gov/Publications/ERR43/
41. Just DR. Behavioral economics, food assistance, and obesity. Agr Resource Econ Rev. 2006;35(2):209–20.
42. Wilcox K, Vallen B, Block L, Fitzsimons GJ. Vicarious goal fulfillment: when the mere presence of a
healthy option leads to an ironically indulgent decision. J Consum Res. 2009;36(3):380–93.
Food Choices
HEALTH AFFAIRS ~ Web Exclusive w1121