How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity

Abstract

This paper uses data on daily activities from the American Time Use Survey and the associated Eating & Health Module to analyze
the relationships between time poverty and specific energy-balance behaviors. The authors estimate a simultaneous model to
jointly analyze the relationships between time poverty and the probability of a fast food purchase, the number of eating and
drinking occurrences, minutes spent engaging in sports and exercise, and the probability of engaging in active travel (walking
or cycling). Time-poor individuals were found to have different eating and physical activity patterns than non-time-poor individuals;
those who were time-poor were less likely to purchase fast food and also less likely to engage in active travel.

Full text

Submitted Article
How Does Time Poverty Affect Behavior? A
Look at Eating and Physical Activity
Charlene M. Kalenkoski and Karen S. Hamrick*
Charlene M. Kalenkoski, Department of Economics, Ohio University. Karen
S. Hamrick, Economic Research Service, U.S. Department of Agriculture. The
views expressed here are those of the authors and are not necessarily those of the
Economic Research Service or the U.S. Department of Agriculture.
Editor: Ted McDonald
*Correspondence to be sent to: khamrick@ers.usda.gov.
Submitted 15 December 2011; accepted 24 August 2012.
Abstract This paper uses data on daily activities from the American Time Use
Survey and the associated Eating & Health Module to analyze the relationships
between time poverty and specific energy-balance behaviors. The authors estimate
a simultaneous model to jointly analyze the relationships between time poverty
and the probability of a fast food purchase, the number of eating and drinking
occurrences, minutes spent engaging in sports and exercise, and the probability of
engaging in active travel (walking or cycling). Time-poor individuals were found
to have different eating and physical activity patterns than non-time-poor individ-
uals; those who were time-poor were less likely to purchase fast food and also less
likely to engage in active travel.
Key words: Time use, Discretionary time, Time poverty, Time-poor,
American Time Use Survey, Eating and Health Module, Energy balance,
Exercise, Eating patterns, Fast food, Active travel.
JEL codes: I12, J10, I30.
Introduction
Time poverty is defined as not having enough discretionary time.
Discretionary time is important for restorative purposes and for invest-
ment in one’s health and human capital; it is also important for avoiding
social exclusion (Bittman 2002). This study defines an individual’s daily
discretionary minutes by subtracting minutes of necessary and committed
time from 1,440 (total daily minutes) as done in Kalenkoski, Hamrick, and
Andrews (2011). Necessary activities are those activities that must be per-
formed by an individual for him- or herself (sleep, grooming, health-
related self-care, and other personal and/or private activities). Committed
activities are those that must be performed due to previous life choices
such as whether to marry, to divorce, to have and raise children, and to be
employed (time spent in household work, time spent in child care, time
Published by Oxford University Press on behalf of Agricultural and Applied Economics Association 2012.
Applied Economic Perspectives and Policy (2013) volume 35, number 1, pp. 89– 105.
doi:10.1093/aepp/pps034
Advance Access publication on 14 November 2012.
89
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

spent caring for household adults, and time spent in employment or
related activities).
1
Although an individual has the ability to make mar-
riage, divorce, fertility, and employment choices over time, on a given day
these choices have already been made, and their associated time commit-
ments are thus essentially fixed. Therefore, although each individual tech-
nically has 24 hours in a day, different people face different discretionary
time constraints depending on their life circumstances.
This study hypothesizes that time-poor individuals may not be able to
prepare and eat healthy meals or to exercise. These hypotheses are tested
by estimating a simultaneous model that jointly analyzes the relations
between time poverty and fast food purchases, the number of eating and
drinking occurrences, minutes spent engaging in sports and exercise, and
engaging in active travel (walking or biking twenty minutes or more).
Better understanding of Americans’ eating and activity behaviors can
provide insight into policies and programs that address obesity issues.
Theory
In the standard labor-leisure model, discretionary time, T, is divided
into two categories, labor and leisure. Paid work would fall under the cat-
egory of labor, while preparing and eating healthy meals and exercising
would fall under the category of leisure. While predictions from the
model based on T can be made, most analyses assume that T is fixed and
is the same for all individuals. However, due to people’s differing com-
mitments, for example, due to a minimum amount of child care that must
be performed, people do indeed have different levels of discretionary
time. In figure 1, individual 1 has more discretionary time, T
1
, than indi-
vidual 2, T
2
. Because the individuals’ discretionary time is different, the
maximum hours that they work and the maximum income from their
work are different, even if their wage rates are the same. Consequently,
individual 1, with greater discretionary time, will be able to attain a
labor-leisure choice on a higher indifference curve with greater utility (U
1
)
than individual 2 (U
2
). Labor-leisure choice A on U
1
allows for more
hours worked and more income (I
1
), but also more discretionary hours
(D
1
) than individual 2’s choice B with less income (I
2
) and less discretion-
ary time (D
2
).
Literature Review
The concept of time poverty is not new. Vickery (1977),Douthitt (2000),
and Davis and You (2011) modified existing income poverty thresholds to
account for time constraints. Gershuny (2011) developed the “triangle of
daily activities,” which includes leisure, unpaid work, and paid work to
represent work-life balance when measuring national well-being. Other
authors defined time poverty and calculated time poverty thresholds for
various countries: Bardasi and Wodon (2006) for Guinea; Harvey and
Mukhopadhyay (2007) for Canada; McGinnity and Russell (2007) for
Ireland; Burchardt (2008) for the United Kingdom; Spinney and Millward
1
More detail regarding the classification of necessary, committed, and discretionary activities is pro-
vided in the appendix.
Applied Economic Perspectives and Policy
90
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

(2010) for Canada; and Kalenkoski, Hamrick, and Andrews (2011) for the
United States. Zacharias (2011) compared alternative approaches to defin-
ing time and income poverty and discussed their potential role in anti-
poverty policies.
Fewer studies, however, have investigated the associations of time
poverty and individuals’ behavior. Mothersbaugh, Hermann, and Warland
(1993) examined the relationship between perceived time pressure and
people’s adherence to recommended dietary practices (RDPs), and found
that perceptions of time pressure do indeed have adverse effects on indi-
viduals’ eating habits and RDPs. Although he did not directly measure
time poverty, Christian (2009) found that longer commutes were associated
with less time spent in exercise and other health-related activities, and were
also associated with substitution into lower-intensity exercise. In addition,
he found that longer commutes increase the likelihood of non-grocery food
purchases. Spinney and Millward examined the associations between time
and income poverty and participation in moderate or higher intensity phys-
ical activities, and concluded that “ ... time poverty may be more important
than income poverty as a barrier to regular physical activity,” (Spinney and
Millward 2010).
Eating patterns have been found to matter for health outcomes. By
using the Seasonal Variation of Blood Cholesterol Study data, Ma et al.
(2003) found that a greater number of eating episodes per day was associ-
ated with a lower risk of obesity. Using the American Time Use Survey
(ATUS) Eating and Health (EH) Module data, Hamermesh (2010) found a
similar result, that the frequency of eating meals (primary eating occur-
rences) and grazing (secondary eating occurrences) were associated with
lower BMI and better self-reported health. Kolodinsky and Goldstein
(2011) also used the ATUS and EH Module data to investigate the relation-
ships between time use and food patterns and obesity, and found that
Figure 1 Labor-Leisure Choices with Different Amounts of Discretionary Time
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
91
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

increases in time spent in meal preparation and cleanup are associated
with decreases in BMI.
Eating food away from home (FAFH) also has been found to matter for
health outcomes. Binkley, Eales, and Jekanowski used the 1994-96
Continuing Survey of Food Intake by Individuals (CSFII) and found a pos-
itive relationship between respondents’ BMI and FAFH consumed in the
previous 24 hours. They concluded that “FAFH, and particularly fast food
consumption, are likely to be contributing factors to increased obesity,”
(Binkley, Eales, and Jekanowski 2000). Todd, Mancino, and Lin, using the
1994-96 CSFII and the 2003-04 National Health and Nutrition Examination
Survey (NHANES) data, concluded that FAFH “ ...is a contributing factor
to poor diet quality and that concern about FAFH’s effect on obesity is
warranted,” (Todd, Mancino, and Lin 2010). These authors also found that
one additional meal eaten away from home increases daily caloric intact
by about 134 calories, and lowers diet quality by two points on the
Healthy Eating Index.
Looking specifically at fast food, the literature is mixed on the effects of
fast food on obesity. Binkley et al. (2000) and Jeffery et al. (2006) found
that fast food consumption was positively associated with BMI. Mehta
and Chang (2008) studied food environment and found that a higher ratio
of fast food to full-service restaurants was associated with higher BMIs.
Bowman and Vinyard (2004) found an association between fast food con-
sumption and overweight status, but characterized that relationship as
being weak. However, Cutler et al. (2003) found that Americans’ increased
caloric intake is from more snacks, whereas the increase in fast food con-
sumption is a result of reduced home consumption. Thus, fast food calo-
ries are offsetting fewer calories eaten at home. Despite this mixed
evidence, the Institute of Medicine of the National Academies report
(2012) categorizes fast food restaurants as “unhealthy food venues” (p. 3).
Physical activity also affects health outcomes. Hemmingsson and
Ekelund (2007) collected their own data measuring physical activity with
accelerometry, and tested the generally-accepted inverse relationship
between physical activity and BMI. They found that physical activity and
BMI are only weakly associated for non-obese individuals, but are highly
significantly associated for obese individuals. Stamatakis, Hirani, and
Rennie (2009) investigated the relationships of physical activity types—
from inactive to sufficiently active for obesity prevention—and sedentary
behavior with BMI using the 2003 Scottish Health Survey. These authors
concluded that physical activity and sedentary behavior are both strongly
and independently related to obesity, for obesity defined as BMI ≥30 kg/
m
2
and waist circumference (WC) ≥88 cm in women and ≥102 cm in
men. Dunton et al. used the 2006 ATUS and EH Module data to examine
the interaction between time spent in physical activity and in sedentary
behavior on BMI. They concluded that “sedentary behaviors and physical
activity interact with each other in relation to BMI in adults,” (Dunton
et al. 2009).
Long-run trends in obesity have also been studied (Cutler et al. 2003;
Variyam 2005), and a variety of environmental factors have been identified
as contributing to these long-run trends. Powell and Chaloupka (2009)
point to the decline in food prices and the increase in the availability of fast
Applied Economic Perspectives and Policy
92
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

food and restaurant meals, and Chou et al. (2004) point to the increased
price of cigarettes and the reduction in smoking. Another important con-
tributor to the increase in Americans’ body size has been the changing
nature of work. Both technological advances and industrial restructuring in
the United States have resulted in the decline in physical exertion required
for performing paid work. Philipson and Posner stated this trend concisely
by noting that in an agricultural or industrial society, “work is strenuous;
in effect, the worker is paid to exercise,” whereas in a post-industrial
society, “people must pay for undertaking—rather than be paid to under-
take—physical activity” (Philipson and Posner 2003). As a result, individu-
als then pay for exercise by budgeting their recreational time. “The full
price of physical activity is the opportunity cost of allotted time—the value
of the most preferable alternative given up by allotting time for a walk in
the neighborhood or a run in the park” (Variyam 2005).
The contribution of the present study is to utilize the very detailed and
comprehensive ATUS data to examine both sides of the energy-balance
equation–both input and expenditure of energy. As such, it analyzes how
time poverty is associated with fast food purchases, the number of eating
and drinking occurrences, the number of minutes engaged in sports and
other exercise, and time engaged in active travel.
Data
The Bureau of Labor Statistics’ American Time Use Survey data were
used for this study. One individual aged 15 or older from each sampled
household was interviewed about his or her activities for the 24-hour
period from 4 A.M.the day before the interview to 4 A.M.the day of the
interview. Survey respondents were asked to identify their primary activ-
ity if they were engaged in more than one activity at a time. They were
also asked to report where they were and who else was present for each
activity. In addition to the time-diary data, demographic, labor force par-
ticipation, and household information was collected from the respondents.
This study used the Respondent, Activity, Activity Summary, and
Methodology (Case History) files from the ATUS, as well as the EH
Module Respondent, Activity, and Replicate Weights files. The analysis
was limited to the years 2006-2008 because the EH files were available for
this period only.
2
From 2006-08, the ATUS and EH Module resulted in
37,832 completed interviews of individuals aged 15 or over. The replicate
weights from the EH Module produce nationally-representative estimates
for an average day over this period. Excluding those with bad diaries,
3
2
See U.S. DOL BLS (2010) for discussion of using the American Time Use Survey data, and Hamrick
(2010) and Hamrick et al. (2011) for discussion of using the Eating & Health Module data.
3
After each ATUS interview is completed, the Census interviewer answers two data quality questions:
“Is there any reason the information from this interview should NOT be used?” and “Why do you
think the data should not be used?” In 275 cases of the 37,832 completed interviews, the Census inter-
viewer thought that the respondent’s time diary was not of good quality, as indicated by the variable
TUDQUAL2 from ATUS Case History data file. We defined a poor-quality time diary as one where
TUDQUAL2 had a value of 1 (intentionally wrong), 2 (could not remember), 3 (deliberately long
durations), or 4 (other reason).
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
93
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

those under age 20, and those that are underweight,
4
the resulting sample
size is 32,392.
5
One potential drawback of the ATUS diary data is that information on
only one time-diary day per person was collected. There may be concern
that some activities, such as eating fast food or engaging in sports and
exercise, are not daily activities and thus that a one-day diary such as the
ATUS lacks intrapersonal variability. However, some activities, such as
eating patterns, have a large degree of persistency, meaning that
day-to-day variation is minimal; Wansink’s (2007) Mindless Eating dis-
cusses the myriad external influences that result in eating habits. Exercise
is also considered to be a habit, and researchers have studied what con-
tributes to habitual exercise (Aarts et al. 1997;Finlay et al. 2002). Indeed,
much of an individual’s daily activities can be classified as habitual repeti-
tion (Neal et. al. 2006).
Nevertheless, food intake surveys are typically multiday surveys. For
example, the NHANES includes two 1-day food recall interviews.
6
However, because the second-day diaries have a higher rate of non-
response, and because respondents’ consistent reports of less food con-
sumption on the second day suggest under-reporting, some researchers
elect to use only the first diary day (Gregory et al. 2012).
Indeed, existing research supports using a one-day diary to analyze
individuals’ activity patterns. Lambe et al. (2000) examined food consump-
tion using 14-day diaries in five locations in the European Union. Among
their findings was that the quality of the diaries declined over the 14 days,
with the best information and most variation obtained in the first three
days. However, they found that mean intakes of a given food item were
not affected by survey duration. More recently, Raux et al. (2011) studied
seven-day travel diaries for individuals in Ghent, Belgium, and concluded
that while there is a large amount of interpersonal variability (differences
across individuals in their travel patterns), there is small intrapersonal var-
iability (variation across an individual’s seven days of time diaries).
Likewise, Schmidt (2011) studied seven-day diaries of Germans’ payments
(consumer expenditures), both cash and noncash, with a focus on cash
payments, and found that survey fatigue is apparent and that more cash
payments were recorded on day one. However, the distribution of pay-
ments during diary days 2–7 is similar, leading Schmidt to conclude “that
additional diary days only increase the sample size, rather than provide
additional information” Schmidt (2011).
4
We exclude those under 20 years old because the Centers for Disease Control adult BMI interpreta-
tion is for persons aged 20 and over. Body Mass Index is calculated as: weight (lb) / [height (in)]
2
×
703. Adult BMI groups are underweight (BMI ,18.5), normal weight (18.5 ≤BMI ,25), overweight
(25 ≤BMI ,30), and obese (30 ≤BMI). See CDC for more information on adult BMI: http://
www.cdc.gov/healthyweight/assessing/bmi/index.html. We exclude those who are underweight because
of the small number of respondents. Those who have a BMI less than 18.5 comprise 1.4% of respond-
ents aged 20 and over with a BMI value. It is the authors’ experience that because of the small cell
size, this group’s characteristics and patterns can be dominated by a small number of respondents who
engage in activities for a long duration. Having a small number of respondents dominate an estimate,
in some cases as few as 2 respondents, is an indicator that the cell size is not large enough for
analysis.
5
Creation of the analysis data set was done using SAS 9.2 and STATA 12 and estimation was per-
formed using STATA 12.
6
For more information on NHANES, see http://www.cdc.gov/nchs/nhanes.htm.
Applied Economic Perspectives and Policy
94
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

Variables
Two measures of eating patterns are available using the ATUS data, an
indicator variable for whether or not fast food was purchased on the diary
day, and the number of eating and drinking occurrences on the diary day
(both primary and secondary).
7
Two measures of physical activity can be
created using the ATUS data, minutes spent on sports and exercise (ATUS
activities 1301xx) on the diary day, and an indicator variable for whether
or not the respondent engaged in active travel (walked or biked twenty
minutes or more on the diary day (ATUS activities 18xxxx and
TEWHERE ¼14 or TEWHERE ¼17). Twenty minutes of daily exercise is
a federal guideline for weight management.
8
Eating and drinking is a daily activity for almost everyone, and food
consumption away from home takes place several times a week, on
average.
9
Active travel is likely to be a daily activity for those who engage
in it, especially if they walk or cycle for all or part of a commute to work.
Therefore, the only measure that is likely to be underestimated by using
one-day diary data is sports and exercise, which is typically not a daily
activity, although it can be considered a habitual activity.
The key explanatory variable in this analysis is an indicator for whether
or not individuals are time-poor. This indicator takes a value of one if
total daily discretionary time is less than 289.8 minutes (4.83 hours), and a
value of zero otherwise. The cutoff of 60% of the median discretionary
time as calculated by Kalenkoski, Hamrick, and Andrews (2011) was
used.
10
Discretionary time is defined as total daily minutes (i.e., 1440)
minus time spent on personal care, market work, household work, child
care, and adult care.
Other explanatory variables available in the ATUS are an indicator vari-
able for whether or not the diary day that was a holiday or weekend day;
an indicator variable for whether or not the respondent is female; age
measured in number of years; an indicator variable for whether or not the
respondent is married; the number of children in the household; an
7
We defined a fast-food purchase as one where travel (180782) was followed by food purchase (070103)
at a restaurant (TEWHERE ¼4). We included both respondents who ate the food at the restaurant
and those who carried out the food. By including only the 070103 cases with TEWHERE ¼4, we
excluded purchasing food at TEWHERE ¼7 (other store, mall) and TEWHERE ¼11 (other place).
The U.S. Census Bureau has identified that secondary eating at these “other place” locations is usually
at an entertainment venue such as a stadium or movie theater (e-mail correspondence from BLS to
Karen S. Hamrick dated February 29, 2008). By selecting those who pay first, that is, those who pur-
chase at a counter-service restaurant, we excluded those who were at sit-down restaurants and reported
talking with waiters and waitresses or interacting with restaurant cashiers at the end of the meal,
which are also coded as 070103. Including these individuals did not qualitatively change the results.
Also note that selecting those cases where the individual paid first will include both fast food or
quick service restaurants (e.g., McDonald’s, Kentucky Fried Chicken, Dunkin’ Donuts) and fast
casual restaurants (e.g., Panera Bread, Cosi, Corner Bakery), which tend to have fresher, lighter fare
than fast food restaurants.
8
http://www.cnpp.usda.gov/Publications/DietaryGuidelines/2010/PolicyDoc/Chapter2.pdf. Page 17.
9
The National Restaurant Association’s 2000 survey found that Americans aged 8 and older consume
an average of 4.2 commercially prepared meals per week (National Restaurant Association 2000). Men
are more likely to consume commercially prepared meals. About 22% of men’s meals, and about 18%
of women’s meals are commercially prepared. Because restaurant sales have increased in real dollars by
about two-thirds from 2000– 2012 (projected), the frequency of consumption of commercially-prepared
meals is likely to have increased since the 2000 survey (National Restaurant Association 2012).
10
Kalenkoski, Hamrick, and Andrews (2011) performed an extensive sensitivity analysis using alterna-
tive definitions of time poverty. The results of those sensitivity analyses are discussed in that paper.
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
95
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

indicator variable for whether income is greater than 185% of the poverty
threshold;
11
an indicator variable for whether income is missing; and an
indicator variable for whether or not the respondent has at least a bache-
lor’s degree.
Model
Three probit models and two continuous regressions are jointly esti-
mated:
F∗=BF0+BF1TP +BF2X+EF
F=1if F∗.0
F=0if F∗≤0
A∗=BA0+BA1TP +BA2X+EA
A=1if A∗.0
A=0if A∗≤0
TP∗=BT0+BT1X+BT2Y+ET
TP =1if TP∗.0
TP =0if TP∗,0
N=BN0+BN1TP +BN2X+EN
S=BS0+BS1TP +BS2X+ES,
where F* is a latent variable representing a fast food purchase; Fis an
indicator variable equal to 1 if fast food was purchased and 0 otherwise;
A* is a latent variable representing active travel; Ais an indicator variable
equal to 1 if the respondent engaged in active travel on the diary day and
0 otherwise; TP* is an indicator variable representing non-discretionary
time; TP is an indicator variable equal to 1 if the respondent was time-
poor and 0 otherwise; Nis the number of eating occurrences on the diary
day; and Sis the number of minutes spent on sports and exercise on the
diary day.
Note that TP is the key explanatory variable in each of the time-use
equations. Because it is a potentially endogenous regressor, it is modeled
jointly with these uses of time. The variable Xis a vector of demographic
and other characteristics of the respondent and his/her household, while
Yis a vector of additional variables in the time poverty probit that may
help identify time poverty in the other equations. The B’s are estimated
coefficients on the explanatory variables, and the E’s are the error terms.
Subscripts for the individual are suppressed for clarity. Because all of the
activities we analyze were engaged in by the same individual, and
because they were all subject to the same time constraint, we allow the
error terms to be correlated across models.
11
The EH Module asks respondents whether their household income is greater or less than the dollar
amount of 185% of the poverty threshold, which corresponds with the income eligibility thresholds for
reduced-price school meals and the Special Supplemental Nutrition Program for Women, Infants, and
Children (WIC). We use the 185% poverty threshold level to define individuals in low-income
households.
Applied Economic Perspectives and Policy
96
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

Results
Table 1presents weighted descriptive statistics for the full sample
(those aged 20 and over who participated in the EH module and also pro-
vided good time diaries), and for the time-poor and not time-poor sub-
samples. These descriptive statistics reflect an average day for the years
2006–08. There are 6,571 persons (20%) in the sample categorized as time-
poor, and 25,821 persons (80%) are categorized as not time-poor. Eight
percent of Americans aged 20 and older purchased fast food on an
average day. Seven percent of those who were time-poor purchased fast
food, while 9% of those who were not time-poor purchased fast food. The
average number of eating and drinking occurrences
12
per day was just
under three, with time-poor individuals having fewer eating and drinking
occurrences (2.7) than not time-poor individuals (2.9). Nineteen minutes
was the daily average number of minutes spent on sports and exercise,
but there was a dramatic difference between the average number of
minutes spent on these activities by those who were time-poor (6 minutes)
and those who were not time-poor (22 minutes). Five percent of respond-
ents were engaged in active travel; 4% of the time-poor and just over 5%
of the not time-poor.
Men were more likely to be time-poor, perhaps because of their greater
labor force participation. Younger individuals were more likely to be time-
poor than older individuals, while married people and people living with
more children in their households were also more likely to be time-poor.
Higher-income individuals were more likely to be time-poor, as were
more highly educated individuals.
Table 2presents results from a maximum likelihood model that jointly
estimates the fast food and active travel probits and the continuous regres-
sions representing the number of eating and drinking occurrences and the
number of minutes spent on sports and exercise. This model accounts for
correlations in the errors of the activity equations but treats time poverty
as exogenous (i.e., excludes the time poverty equation from the model).
Time-poor individuals have different eating patterns than not-time-poor
individuals. Being time-poor is associated with a reduction in the likeli-
hood of a fast food purchase on an average day by 3%, perhaps due to the
time needed to travel to a fast food establishment and to wait in line; that
is, fast food may not in fact be fast to those with limited time. If fast food
is indeed unhealthy food, this is good news for time-poor individuals.
Time poverty is also associated with a reduction in the number of eating
and drinking occurrences on an average day by 0.27. This is a substantial
effect, given that the average number of eating and drinking occurrences
for the full sample is 2.9. According to Ma et al. (2003) and Hamermesh
(2010), this means a greater risk of obesity for time-poor individuals.
Time-poor individuals also have different activity patterns than do
not-time-poor individuals. Time-poor individuals spend almost 18
minutes less on sports and exercise on a given day than do not-time-poor
individuals, a large effect given that the average time spent by all individ-
uals is almost 19 minutes. Similarly, time-poor individuals are 1% less
likely to engage in active travel than not-time-poor individuals. Thus,
12
Eating and drinking occurrences include both primary eating and drinking (eating and drinking
beverages as a main activity) and secondary eating and secondary drinking (eating and/or drink-
ing beverages as a secondary activity while doing something else as the primary activity).
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
97
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

Table 1 Weighted Means and Standard Errors, Means for an Average Day, 2006–08, age 20+
All N 532,392 Time-Poor N 56,571
Not Time-Poor
N525,821
Purchased fast food, 1 ¼YES and 0 ¼NO 0.084 (0.002) 0.066 (0.004)
*,#
0.089 (0.002)
#
Number of eating and drinking occurrences, primary and secondary 2.898 (0.014) 2.747 (0.025)
*,#
2.945 (0.016)
#
Minutes spent on sports and exercise 18.518 (0.434) 6.017 (0.343)
*,#
22.365 (0.553)
*,#
Walked or biked twenty minutes or more, 1 ¼YES and 0 ¼NO 0.050 (0.002) 0.044 (0.003)
#
0.052 (0.002)
Time-Poor, 1 ¼YES and 0 ¼NO 0.235 (0.003) - -
Diary day is weekend or holiday, 1 ¼YES and 1 ¼NO 0.300 (0.001) 0.133 (0.004)
*,#
0.352 (0.002)
*,#
Female, 1 ¼female and 0 ¼male 0.496 (0.001) 0.471 (0.007)
*,#
0.504 (0.002)*
Age (years) 47.097 (0.038) 41.364 (0.181)
*,#
48.861 (0.082)
*,#
Married, 1¼YES and 0 ¼NO 0.594 (0.003) 0.645 (0.007)
*,#
0.579 (0.004)
*,#
Number of household children 0.705 (0.005) 1.013 (0.019)
*,#
0.611 (0.006)
*,#
Income .185% poverty threshold, 1 ¼YES and 0 ¼NO 0.691 (0.003) 0.730 (0.007)
*,#
0.679 (0.004)
#
Income missing, 1 ¼YES and 0 ¼NO 0.027 (0.001) 0.013 (0.002)
*,#
0.031 (0.002)
#
Bachelor’s degree or higher, 1 ¼YES and 0 ¼NO 0.290 (0.003) 0.330 (0.007)
*,#
0.277 (0.004)
*,#
Notes: Standard errors are in parentheses. Eating & Health Module replicate weights were used to calculate the standard errors.
* indicates estimate is significantly different from the total population estimate at the 90% confidence level.
# indicates that Time-Poor and Not Time-Poor estimates are significantly different from each other at the 90% confidence level.
Source: American Time Use Survey and Eating & Health Module data.
Applied Economic Perspectives and Policy
98
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

time-poor individuals are at risk of obesity given that they expend fewer
calories through physical activity than do not-time-poor individuals.
RhoFN is the estimated error correlation between a fast food purchase
and the number of eating and drinking occurrences on an average day.
RhoFN is positive and statistically significant, suggesting that there is an
Table 2 Time Poverty and Daily Eating and Activity Patterns: Time Poverty
Treated as Exogenous, Estimated Marginal Effects, and Standard Errors
Fast Food
Number of
Eating and
Drinking
Occurrences
Minutes
Spent in
Sports and
Exercise
Active
Tr a ve l
Time-poor -0.034*** -0.273*** -17.647*** -0.012***
(0.004) (0.031) (0.778) (0.003)
Female 0.015*** 0.140*** -10.523*** -0.007**
(0.004) (0.027) (0.835) (0.003)
Age -0.001*** 0.002** -0.202*** -0.001***
(0.000) (0.001) (0.030) (0.000)
Married -0.006 0.128*** -1.141 -0.036***
(0.005) (0.031) (0.924) (0.004)
Number of children 0.005** 0.010 -0.502 0.001
(0.002) (0.012) (0.412) (0.002)
Income .185% of income
poverty threshold
0.042*** 0.283*** 8.259*** -0.025***
(0.004) (0.031) (1.006) (0.004)
Income missing -0.022* 0.197 7.851** -0.011
(0.012) (0.173) (3.079) (0.009)
Bachelor’s degree 0.003 0.321*** 4.454*** 0.026***
(0.005) (0.026) (1.022) (0.004)
Weekend or holiday -0.011*** -0.248*** 3.669*** -0.008***
(0.004) (0.029) (1.008) (0.003)
Error Correlations
RhoFN–between fast food purchase and the number of eating
and drinking occurrences
0.045***
(0.011)
RhoFS–between fast food purchase and minutes spent in sports
and exercise
-0.022*
(0.013)
RhoFA–between fast food purchase and engaging in active travel 0.033
(0.031)
RhoNS–between the number of eating and drinking occurrences
and minutes spent in sports and exercise
0.015
(0.010)
RhoNA–between the number of eating and drinking occurrences
and engaging in active travel
0.025*
(0.011)
RhoSA–between minutes spent in sports and exercise and engaging
inactive travel
0.054***
(0.014)
Number of observations 32,392
Prob .Chi2 0.0000
Notes: An intercept is included in all equations. Standard errors are in parentheses and were
calculated using Eating and Health Module replicate weights.
Marginal effects for the fast food and active travel probit equations are calculated at the mean. For
discrete explanatory variables, marginal effects measure the effects of discrete changes in the dummy
variables from 0 to 1.
*** indicates statistical significance at the 1% level, ** indicates statistical significance at the 5% level,
and * indicates statistical significance at the 10% level.
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
99
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

Table 3 Time Poverty and Daily Eating and Activity Patterns: Time Poverty
Treated as Endogenous Estimated Coefficients, Standard Errors, and Calculated
Marginal Effects
Fast Food
Number of
Eating and
Drinking
Occurrences
Minutes
Spent in
Sports and
Exercise
Active
Tr a ve l
Time-
Poor
Time-poor -0.320*** 2.224 -16.172 -0.179**
(0.066) (1.516) (49.114) (0.090)
[-0.041] [-0.015]
Female 0.103*** 0.174*** -10.503*** -0.074** -0.046
(0.026) (0.036) (1.084) (0.033) (0.098)
[0.015] [-0.007] [-0.014]
Age -0.010*** 0.012* -0.196 -0.009*** -0.013***
(0.001) (0.006) (0.199) (0.001) (0.004)
[-0.001] [-0.001] [-0.004]
Married -0.036 0.025 -1.202 -0.360*** 0.129***
(0.032) (0.073) (2.254) (0.036) (0.033)
[-0.005] [-0.035] [0.040]
Number of children 0.035*** -0.079 -0.555 0.013 0.093***
(0.014) (0.055) (1.823) (0.018) (0.022)
[0.005] [0.001] [0.029]
Income .185% of
income poverty
threshold
0.318*** 0.220*** 8.221*** -0.253*** 0.089
(0.033) (0.048) (1.600) (0.040) (0.070)
[0.043] [-0.025] [0.027]
Income missing -0.171* 0.316* 7.921* -0.131 -0.295***
(0.103) (0.185) (4.263) (0.115) (0.113)
[-0.022] [-0.011] [-0.082]
Bachelor’s degree 0.020 0.253*** 4.414** 0.257*** 0.091***
(0.035) (0.056) (1.770) (0.038) (0.035)
[0.003] [0.026] [0.029]
Weekend or holiday -0.087*** 0.220 3.945 -0.095*** -0.581***
(0.030) (0.283) (9.218) (0.037) (0.094)
[-0.012] [-0.008] [-0.165]
Error Correlations
RhoFN–between fast food purchase and the number of
eating and drinking occurrences
0.027
(0.019)
RhoFS–between fast food purchase and minutes spent in
sports and exercise
-0.023*
(0.014)
RhoFA–between fast food purchase and engaging in active
travel
0.032
(0.030)
RhoFT–between fast food purchase and time poverty 0.036
(0.035)
RhoNS–between the number of eating and drinking
occurrences and minutes spent in sports and exercise
0.018
(0.018)
RhoNA–between the number of eating and drinking
occurrences and engaging in active travel
0.011
(0.019)
RhoNT–between the number of eating and drinking
occurrences and time poverty
-1.058
(0.647)
RhoSA–between the minutes spent in sports and exercise
and engaging in active travel
0.054***
(0.014)
RhoST–between the minutes spent in sports and exercise
and time poverty
-0.016
(1.167)
Continued
Applied Economic Perspectives and Policy
100
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

unobserved factor, perhaps a time constraint, that limits long periods for
meals and thus leads to a greater probability of a fast food purchase and
more short eating occurrences throughout the day. RhoFS is the estimated
error correlation between a fast food purchase and the number of minutes
spent on sports and exercise; it is negative and statistically significant, sug-
gesting that an unobserved factor, such as a preference for a healthy life-
style, leads to a lower likelihood of a fast food purchase and more minutes
spent on sports and exercise. RhoNA, the estimated error correlation
between the number of eating and drinking occurrences and active travel,
is positive and statistically significant, perhaps also reflecting an unob-
served preference for healthy living. Finally, RhoSA is the estimated error
correlation between the number of minutes spent on sports and exercise,
and the probability of active travel; it is positive and statistically signifi-
cant, again suggesting that a person’s desire to have a healthy lifestyle
leads to more time spent on sports and exercise and a greater likelihood of
walking or biking more than 20 minutes on an average day.
Table 3presents results that control for the possibility that time poverty
is endogenous
13
(i.e., a time poverty probit is added to the previous simul-
taneous equations model). Time poverty may be endogenous because it
depends on time spent in other activities such as personal care, market
work, and household work. Although we have argued that these daily
activities are largely fixed given prior commitments to work and family,
there may be discretionary time spent on these activities that was jointly
chosen with time spent eating and engaging in physical activity.
Once the potential endogeneity of time poverty is controlled for, time
poverty is no longer a statistically significant determinant of the number
of eating and drinking occurrences, or the number of minutes spent on
sports and exercise. However, time-poor individuals are still significantly
less likely to purchase fast food, with time-poor individuals 4% less likely
to make a fast food purchase than not-time-poor individuals. They are
also 2% less likely to engage in active travel. These effects are only slightly
larger in magnitude than those from the model that did not account for
Table 3 Continued
Error Correlations
RhoAT– between engaging in active travel and time poverty 0.026
(0.055)
Number of observations 32,392
Prob .Chi2 0.0000
Notes: An intercept is also included in all equations.
Eating and Health Module replicate weights were used to calculate standard errors.
Marginal effects for the fast food, active travel, and time poverty probit equations are calculated at the
mean. For discrete explanatory variables, marginal effects measure the effects of discrete changes in the
dummy variables from 0 to 1.
*** indicates statistical significance at the 1% level, ** indicates statistical significance at the 5% level,
and * indicates statistical significance at the 10% level.
13
The results in tables 2and 3are presented differently. In table 2, marginal effects and standard
errors of the marginal effects are presented for the probit models. However, the complexity of the model
in table 3renders calculation of the standard errors of the marginal effects of the probit models inap-
propriate. Indeed, STATA SE 12 refuses to compute them. Thus, in table 3, for the probit outcomes,
coefficient estimates, standard errors of the coefficient estimates, and marginal effects are presented.
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
101
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

endogeneity. Most of the other explanatory variables retain similar effects
across models. Thus, controlling for endogeneity appears only to eliminate
the associations of time poverty with the number of eating and drinking
occurrences and minutes spent on sports and exercise.
It should be noted that no “instruments” are included in the time
poverty equation to identify time poverty in the other equations. Metro
status, the state-level monthly unemployment rate, and the state-level cost
of child care as a percentage of income were all tried as possible instru-
ments, as each could potentially affect an individual’s discretionary time.
However, none of these was a significant correlate of time poverty. It is
likely that this lack of significance is due to the limited geographic infor-
mation available in the ATUS, which provides information on the state in
which a respondent lives, but these variables are likely more relevant at
the county or local level. Nevertheless, our model is identified by non-
linearities (Roodman 2011).
Conclusion
Understanding the complexities of Americans’ eating and activity
behaviors is important for addressing America’s high obesity rate. These
findings on the relationships between time poverty and fast food pur-
chases, the number of eating and drinking occurrences, time spent
engaged in sports and exercise, and active travel shed light on this issue.
To their benefit, time-poor individuals are less likely to purchase fast
food, which is considered by some to be generally unhealthy. This finding
presents the possibility that time-poor individuals are purchasing pre-
pared food from grocery stores or other venues. Research is needed to
better understand convenience foods and their sources. To their detriment,
time-poor individuals are also less likely to engage in active travel. One
possible policy implication of this finding is that policy-makers could con-
sider the development of sidewalks and bike paths to encourage individu-
als to walk or cycle more frequently. Employers could also implement
programs to encourage active travel to work, such as taking steps to earn
the designation of “Bicycle Friendly Business.”
Acknowledgements
The authors thank John Cawley, Catrine Tudor-Locke, the Editor, and two anony-
mous referees for their comments and suggestions.
References
Aarts, H., T. Paulussen, and H. Schaalma. 1997. Physical Exercise habit: On the
Conceptualization and Formation of Habitual Health Behaviours. Health
Education Research 12:363– 74.
Bardasi, E., and Q. Wodon. 2006. Measuring Time Poverty and Analyzing Its
Determinants: Concepts and Application to Guinea. In C. Mark Blackden,
and Q. Wodon, eds. Gender, Time Use, and Poverty in Sub-Saharan Africa. World
Bank Working Paper No. 73. Washington, DC.
Applied Economic Perspectives and Policy
102
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

Binkley, J. K., J. Eales, and M. Jekanowski. 2000. The Relation Between Dietary
Change and Rising US Obesity. International Journal of Obesity 24(8):1032– 1039.
Bittman, M. 2002. Social Participation and Family Welfare: The Money and Time
Costs of Leisure in Australia. Social Policy and Administration 36(4):408– 425.
Bowman, S. A., and B. T. Vinyard. 2004. Fast Food Consumption of U.S. Adults:
Impact on Energy and Nutrient Intakes and Overweight Status. Journal of the
American College of Nutrition 23(2):163–168.
Burchardt, T. 2008. Time and Income Poverty. Case Report 57. York, UK: Centre for
Analysis of Social Exclusion.
Chou, S.-Y., M. Grossman, and H. Saffer. 2004. An Economic Analysis of Adult
Obesity: Results from the Behavioral Risk Factor Surveillance System. Journal of
Health Economics 23(3):565–587.
Christian, T.J. 2009. Opportunity Costs Surrounding Exercise and Dietary
Behaviors: Quantifying Trade-offs Between Commuting Time and
Health-Related Activities. Working paper, Department of Economics, Georgia
State University, 21 October 2009.
Cutler, D.M., E.L. Glaeser, and J.M. Shapiro. 2003. Why Have Americans Become
More Obese? Journal of Economic Perspectives 17(3):93– 118.
Davis, George C., and W. You. 2011. Not Enough Money or Not Enough Time to
Satisfy the Thrifty Food Plan? A Cost Difference Approach for Estimating a
Money-Time Threshold. Food Policy 36:101–107.
Douthitt, R.A. 2000. ‘Time to Do the Chores?’ Factoring Home-Production Needs
into Measures of Poverty. Journal of Family and Economic Issues 21(1):7 – 22.
Dunton, G.F., D. Berrigan, R. Ballard-Barbash, B. Graubard, and A.A. Atienza.
2009. Joint Associations of Physical Activity and Sedentary Behaviors with Body
Mass Index: Results from a Time Use Survey of US Adults. International Journal
of Obesity 33(12):1427–1436.
Ebbin, R. 2000. Americans’ Dining-Out Habits. Restaurants USA, November issue.
National Restaurant Association. Available at: http://www.restaurant.org/
tools/magazines/rusa/magArchive/year/article/?ArticleID=138. Accessed on:
April 25, 2012.
Finlay, K.A., D. Trafimow, and A. Villarreal. 2002. Predicting Exercise and Health
Behavioral Intentions: Attitudes, Subjective Norms, and Other Behavioral
Determinants. Journal of Applied Social Psychology 32:342–56.
Gershuny, J. 2011. Time-Use Surveys and the Measurement of National Well-being.
Centre for Time Use Research, University of Oxford and Office for National
Statistics. Report dated 12 September 2011.
Gregory, C.A., M. Ver Ploeg, M. Andrews, and A. Coleman-Jensen. 2012. SNAP
and Diet Quality: An Instrumental Variables Approach. Working paper dated June
19, 2012, available from author, cgregory@ers.usda.gov.
Hamermesh, D.S. 2010. Incentives, time use and BMI: The roles of eating, grazing
and goods. Economics & Human Biology 8(1):2–15.
Hamrick, K. 2010. Eating & Health Module User’s Guide (2010 Edition). Washington
DC: U.S. Department of Agriculture, Economic Research Service, Administrative
Publication AP-047A, April 2010.
Hamrick, K.S., M. Andrews, J. Guthrie, D. Hopkins, and K. McClelland. 2011. How
Much Time Do Americans Spend on Food? Washington DC: U.S. Department of
Agriculture Economic Research Service, Economic Information Bulletin EIB-86,
Nov. 2011.
Harvey, A.S., and A.K. Mukhopadhyay. 2007. When Twenty-Four Hours is Not
Enough: Time Poverty of Working Parents. Social Indicators Research 82:57–77.
Hemmingsson, E., and U. Ekelund. 2007. Is the Association Between Physical
Activity and Body Mass Index Obesity Dependent? International Journal of
Obesity 31(4):663–668.
Institude of Medicine of the National Academies. 2012. Accelerating Progress in
Obesity Prevention: Solving the Weight of the Nation. Washington, DC: The
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
103
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

National Academies Press. Retrieved from http://www.iom.edu/~/media/
Files/Report%20Files/2012/APOP/APOP_insert.pdf
Jeffery, R., J. Baxter, M. McGuire, and J. Linde. 2006. Are Fast Food Restaurants
an Environmental Risk Factor for Obesity? International Journal of Behavioral
Nutrition and Physical Activity 3(2). http://www.ijbnpa.org/content/3/1/35;
http://www.ijbnpa.org/content/3/October/2006
Kalenkoski, C.M., K.S. Hamrick, and M. Andrews. 2011. Time Poverty Thresholds
and Rates for the US Population. Social Indicators Research 104(1):129–155.
Kolodinsky, J.M., and A.B. Goldstein. 2011. Time Use and Food Pattern Influences
on Obesity. Obesity, 19(12):2327– 2335.
Lambe, J., et al. 2000. The Influence of Survey Duration on Estimates of Food
Intakes and Its Relevance for Public Health Nutrition and Food Safety Issues.
European Journal of Clinical Nutrition 43(2):166–173.
Ma, Y., et al. 2003. Association between Eating Patterns and Obesity in a
Free-living US Adult Population. American Journal of Epidemiology 158(1):85– 92.
McGinnity, F., and H. Russell. 2007. Work Rich, Time Poor? Time-Use of Women
and Men in Ireland. The Economic and Social Review 38(3):323– 354.
Mehta, N.K., and V.W. Chang 2008. Weight Status and Restaurant Availability.
American Journal of Preventive Medicine 34(2):127– 133.
Mothersbaugh, D.L., R.O. Hermann, and R.H. Warland. 1993. Perceived Time
Pressure and Recommended Dietary Practices: The Moderating Effect of
Knowledge of Nutrition. Journal of Consumer Affairs 27(1):106– 26.
National Restaurant Association. 2012. 2012 Restaurant Industry Pocket Factbook,
National Restaurant Association. Available at: http://www.restaurant.org/pdfs/
research/PocketFactbook_2012.pdf. Accessed on: April 25, 2012.
Neal, D., W. Wood, and J. M. Quinn. 2006. Habits–A Repeat Performance. Current
Directions in Psychological Science 15(4):198– 202.
Philipson, T.J., and R.A. Posner. 2003. The Long-Run Growth in Obesity as a
Function of Technological Change. Perspectives in Biology and Medicine 46(3,
supplement):S87–S107.
Powell, L.M., and F.J. Chaloupka. 2009. Food Prices and Obesity: Evidence and
Policy Implications for Taxes and Subsidies. The Milbank Quarterly
87(1):229–257.
Raux, C., T.Y. Ma, and E. Cornelis. 2011. Variability versus Stability in Daily
Travel and Activity Behaviour: The Case of a One-week Travel Diary. HAL
(Hyper Articles en Ligne). http://halshs.archives-ouvertes.fr/docs/00/61/26/10/
PDF/Raux_et_al_Variability_vs_stability.pdf;http://halshs.archives-ouvertes.fr/
halshs-00612610/en/
Roodman, D. 2011. Fitting Fully Observed Recursive Mixed-Process Models with
cmp (Working Paper 168). Stata Journal 11(2):159– 206.
Schmidt, T. 2011. Fatigue in Payment Diaries– Empirical Evidence from Germany.
Deutsche Bundesbank Discussion Paper Series 1: Economic Studies No. 11/2011.
Frankfurt, Germany.
Spinney, J., and H. Millward. 2010. Time and Money: A New Look at Poverty and
the Barriers to Physical Activity in Canada. Social Indicators Research
99(2):341–356.
Stamatakis, E., V. Hirani, and K. Rennie. 2009. Moderate-to-Vigorous Physical
Activity and Sedentary Behaviours in Relation to Body Mass Index-Defined and
Waist Circumference-Defined Obesity. British Journal of Nutrition 101(05):765– 773.
Todd, J.E., L. Mancino, and B-H. Lin. 2010. The Impact of Food Away From Home on
Adult Diet Quality. Washington DC: U.S. Department of Agriculture, Economic
Research Service. Economic Research Report ERR-90, February 2010.
U.S. Department of Labor, Bureau of Labor Statistics. 2010. American Time Use
Survey User’s Guide: Understanding ATUS 2003 to 2009. Accessed on: July 23,
2010.
Variyam, J.N. 2005. Economics and the Rise in Obesity. Amber Waves 3(1):20–27.
Applied Economic Perspectives and Policy
104
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from

Vickery, C. 1977. The Time-Poor: A New Look at Poverty. Journal of Human
Resources 12(1):27– 48.
Wansink, B. 2007. Mindless Eating: Why We Eat More Than We Think. New York:
Bantam Books.
Zacharias, A. 2011. The Measurement of Time and Income Poverty. Levy
Economics Institute of Bard College, Working Paper No. 690.
Appendix—Necessary, Committed, and Discretionary
Activities by Major Activity Group
ATUS Major
Activity Code
Necessary activities
Personal care (includes sleeping and grooming) 01
Committed activities
Household activities (includes housework, food & drink prep.) 02
Caring for and helping household members, both children and
adults
03
Work and work-related activities 05
Discretionary activities
Caring for and helping non-household members 04
Education 06
Consumer purchases 07
Professional and personal care services (includes banking,
paying for daycare, doctor’s appointment, getting a haircut)
08
Household services (includes dropping off/picking up clothes
from dry cleaner, hiring a plumber for home repair, waiting
while car is repaired)
09
Government services and civic obligations (includes using
social services, getting car inspected, serving on jury duty,
voting)
10
Eating and drinking 11
Socializing, relaxing, and leisure (includes entertaining family
and friends, watching television, computer use for leisure,
attending performing arts event, gambling)
12
Sports, exercise, and recreation (includes participating in sports
and attending a sporting event)
13
Religious and spiritual activities 14
Volunteer activities 15
Telephone calls 16
Note: Related waiting and travel times are included in each use of time. Source: Kalenkoski, Hamrick,
and Andrews (2011).
How Does Time Poverty Affect Behavior? A Look at Eating and Physical Activity
105
at DigiTop USDA's Digital Desktop Library on February 12, 2013http://aepp.oxfordjournals.org/Downloaded from