National time accounting: The currency of life

Tags: activities, N. Schwarz, D. Kahneman, time allocation, D. Schkade, self-reported, National Accounts, cluster analysis, longitudinal studies, self-report, National Income, Arthur A. Stone, Norbert Schwarz, A. B. Krueger, experimental studies, activity, Gross National Product, National Time Accounting, Alan B. Krueger, David Schkade, national income accounting, Robert Kennedy, emotional experiences, leisure activities, Rady School of Management, Stony Brook University, Robert F. Kennedy, National Income Accounts, Princeton University, religious activities, 24 A. B. Krueger, experimental induction, representative sample, immune system, studies show, longitudinal studies show, Schwarz, A. A. Stone, investment activities, Daniel Kahneman, domestic activities, affective states, naturalistic observation, Stone, National Bureau of Economic Research
Content: This PDF is a selection from a published volume from the National Bureau of Economic Research Volume Title: Measuring the Subjective Well-Being of Nations: National Accounts of Time Use and Well-Being Volume Author/Editor: Alan B. Krueger, editor Volume Publisher: University of Chicago Press Volume ISBN: 0-226-45456-8 Volume URL: http://www.nber.org/books/krue08-1 Conference Date: December 7-8, 2007 Publication Date: October 2009 Title: National Time Accounting: The Currency of Life Author: Alan B. Krueger, Daniel Kahneman, David Schkade, Norbert Schwarz, Arthur A. Stone URL: http://www.nber.org/chapters/c5053
1 National Time Accounting The Currency of Life Alan B. Krueger, Daniel Kahneman, David Schkade, Norbert Schwarz, and Arthur A. Stone Time is the coin of your life. It is the only coin you have, and only you can determine how it will be spent. Be careful lest you let other people spend it for you. -- Carl Sandburg 1.1 Introduction The development of the National Income and Product Accounts (NIPA) was arguably the foremost contribution of economics in the last century, and the National Bureau of Economic Research's role in developing the accounts remains an unparalleled achievement. Nearly every country tracks its national income today, and limiting fluctuations in national income is a goal of public policy around the world. The National Accounts have been used to estimate bottlenecks in the economy, to forecast business growth, and to inform government budgeting.1 As then-Treasury Secretary Robert Rubin said, "the development of the GDP measure by the Department of Alan B. Krueger is the Bendheim Professor of Economics and Public Policy at Princeton University. Daniel Kahneman is a senior scholar and professor of psychology and public affairs emeritus at the Woodrow Wilson School of Public and International Affairs, and the Eugene Higgins Professor of Psychology Emeritus, Princeton University. David Schkade holds the Jerome S. Katzin Endowed Chair and is associate dean and a professor of management at the Rady School of Management, University of California, San Diego. Norbert Schwarz is the Charles Horton Cooley Collegiate Professor of Psychology, a professor of business at the Stephen M. Ross School of Business, and research professor at the Institute for Social Research, University of Michigan. Arthur A. Stone is department vice-chair and Distinguished professor of psychiatry and of Psychology at Stony Brook University. We thank the National Institute of Aging, the Hewlett Foundation, and Princeton University for generous financial support. We thank Leandro Carvalho, Marie Connolly, David Kamin, Amy Krilla, Molly McIntosh, and Doug Mills for excellent research assistance, and Ed Freeland, Jack Ludwig, John McNee, and Rajesh Srinivasan for survey assistance. We are grateful to colleagues too numerous to thank individually for their constructive comments and criticisms, but we acknowledge that they have improved our collective U-index. 1. In one important early application, Fogel (2001, 213) describes how Simon Kuznets and Robert Nathan "used national income accounting together with a crude form of linear programming to measure the potential for increased [military] production and the sources from 9
10 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Commerce is a powerful reminder of the important things that government can and does do to make the private economy stronger and our individual lives better."2 Yet gross domestic product (GDP), national income, consumption, and other components of the National Accounts have long been viewed as partial measures of society's well-being--by economists and noneconomists alike. For one thing, the National Accounts miss "near-market" activities, such as home production (e.g., unpaid cleaning, cooking, and child care), which produce services that could be purchased on the market. Perhaps more significantly, the National Accounts do not value social activities, such as interactions between friends or husbands and wives, which have an important effect on subjective well-being. Because economic activity is measured by prices, which are marginal valuations in perfectly competitive markets, the National Accounts miss consumer surplus from market transactions. Diamonds are counted as more valuable than water, for example, yet one could question whether diamonds contribute more to society's well-being. Other limitations of the National Accounts that have long been recognized are: externalities improperly accounted for; prices distorted in imperfectly competitive markets; and the particular distribution of income in a country influences prices and marginal valuations. While attempts have been made to adjust the National Accounts for some of these limitations--such as by valuing some forms of nonmarket activity--these efforts are unlikely to go very far in overcoming these problems. Many of these sentiments were alluded to by Robert Kennedy in his speech "On Gross national product" at the University of Kansas on March 18, 1968: Too much and for too long, we seemed to have surrendered personal excellence and community values in the mere accumulation of material things. Our Gross National Product . . . if we judge the United States of America by that . . . counts Air Pollution and cigarette advertising, and ambulances to clear our highways of carnage. It counts special locks for our doors and the jails for the people who break them. It counts the destruction of the redwood and the loss of our natural wonder in chaotic sprawl. . . . And the television programs which glorify violence in order to sell toys to our children. Yet the Gross National Product does not allow for the health of our children, the quality of their education or the joy of their play. It does not include the beauty of our poetry or the strength of our marriages, the intelligence of our public debate or the integrity of our public officials. It measures neither our wit nor our courage, neither our wisdom nor our learning, neither our compassion nor our devotion which it would come and to identify the materials that were binding constraints on expansion" prior to the U.S. entry in World War II. 2. Quoted from "GDP: One of the Great Inventions of the 20th Century," Survey of Current Business, January 2000.
National Time Accounting: The Currency of Life 11 to our country, it measures everything in short, except that which makes life worthwhile.3 The problem is not so much with the National Accounts themselves as with the fact that policymakers and the public often lose sight of their limitations, or misinterpret national income as the sole object of policy and primary measure of well-being.4 In this volume, we propose an alternative way of measuring society's wellbeing, based on time use and affective (emotional) experience. We call our approach National Time Accounting (NTA). National Time Accounting is a set of methods for measuring, categorizing, comparing, and analyzing the way people spend their time, across countries, over historical time, or between groups of people within a country at a given time. Currently, time use is tracked according to the amount of time spent in various activities--such as traveling, watching television, and working for pay--but the evaluation and grouping of those activities is decided by external researchers and coders. Determining whether people are spending their time in more or less enjoyable ways than they were a generation ago is either impossible or subject to researchers' judgments of what constitutes enjoyable leisure activities and arduous work. In addition to the obvious problem that researchers may not view time use in the same way as the general public, other problems with this approach are that: (a) many people derive some pleasure from nonleisure activities; (b) not all leisure activities are equally enjoyable to the average person; (c) the nature of some activities changes over time; (d) people have heterogeneous emotional experiences during the same activities; and (e) emotional responses during activities are not unidimensional. The methods we propose provide a means for evaluating different uses of time based on the population's own evaluations of their emotional experiences, what we call evaluated time use, which can be used to develop a system of national time accounts. We view NTA as a complement to the National Income Accounts (NIA), not a substitute. Like the National Income Accounts, NTA is also incomplete, providing a partial measure of society's well-being. National time accounting misses people's general sense of satisfaction or fulfillment with their lives as a whole, apart from moment to moment feelings.5 Still, we will argue that evaluated time use provides a valuable indicator of society's well-being, and the fact that our measure is connected to time allocation has 3. Transcription available from: www.jfklibrary.org/HistoricalResources/Archives/ ReferenceDesk/Speeches/RFK/RFKSpeech68Mar18UKansas.htm. 4. Kennedy's point has resonance with at least one politician. In an interview, Barack Obama told David Leonhardt (2008) the following: "One of my favorite quotes is--you know that famous Robert F. Kennedy quote about the measure of our G.D.P.? . . . it's one of the most beautiful of his speeches." 5. For surveys of economics research using the more conventional measures of life satisfaction, see Frey and Stutzer (2002) and Layard (2005).
12 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone analytical and policy advantages that are not available from other measures of subjective well-being, such as overall life satisfaction. There have been some attempts at NTA in the past, primarily by time-use researchers. Our approach builds on Juster's (1985) seminal observation that "an important ingredient in the production and distribution of well-being is the set of satisfactions generated by activities themselves" (333). To assess the satisfactions generated by activities, Juster asked respondents to rate on a scale from zero to ten how much they generally enjoy a given type of activity, such as their job or taking care of their children. Later research found that such general enjoyment ratings can deviate in important and theoretically meaningful ways from episodic ratings that pertain to specific instances of the activity (Schwarz, Kahneman, and Xu 2009). To overcome this problem, we utilize a time diary method more closely connected to the recalled emotional experiences of a day's actual events and circumstances. Gershuny and Halpin (1996) and Robinson and Godbey (1997), who analyzed a single well-being measure (extent of enjoyment) and time use collected together in a time diary, are closer forerunners to our approach. Our project is distinguished from past efforts in that we approach NTA from more of a psychological well-being and Experience Sampling Method (ESM) perspective. For example, our measure of emotional experience is multidimensional, reflecting different core affective dimensions. And like ESM, we try to measure the feelings that were experienced during different uses of time as closely as possible. We also developed an easily interpretable and defensible metric of subjective well-being, which combines the data on affective experience and time use to measure the proportion of time spent in an unpleasant state.6 And we use cluster analysis to determine which groups of activities are associated with similar emotional experiences to facilitate the tracking of time use with historical and cross-country data. Past research has not addressed how time-use has shifted among activities associated with different emotional experiences over time, or the extent to which cross-country differences in time allocation can account for international differences in experienced well-being. Lastly, our survey methods attempt to have respondents reinstantiate their day before answering affect questions, to make their actual emotional experiences at the time more vivid and readily accessible for recall. Past calls for National Time Accounting have largely foundered. It is instructive to ask why these efforts were not more influential in academic circles and why government statistical agencies have not implemented them. 6. Because the earlier work focused on whether activities were enjoyable, it would not have been possible to construct our measure of time spent in an unpleasant state with their data. Our approach also differs fundamentally from Glorieux (1993), who asked survey respondents to classify their time use into different "meanings of time," such as social time, time for personal gratification, and meaningless time. Instead, we focus on the emotional experiences that occur over time.
National Time Accounting: The Currency of Life 13 One possible explanation is that it is difficult to collect time diary information along with affective experience in a representative population sample. To this end, we developed a telephone survey, called the Princeton Affect and Time Survey (PATS), patterned on the Bureau of Labor Statistics' (BLS's) American Time Use Survey (ATUS), that is practical and easily adaptable for use in ongoing official time-use surveys. Another possible explanation is that evidence on the validity of subjective well-being measures has progressed greatly in the last decade. While subjective data cannot be independently verified, a range of findings presented in section 1.3 suggests that self-reports of subjective experience indeed have signal. The earlier efforts may have been ahead of their time and taken less seriously than they should have because such evidence was not yet available. Finally, it is difficult to track down documentation on the precise methods used in past diary cum well-being surveys. To facilitate replication and extensions, we have posted our main data sets, questionnaires, and background documents on the web at www.krueger.princeton.edu/Subjective.htm. The remainder of this chapter is organized as follows. Section 1.2 provides a conceptual framework for using evaluated time use in National Time Accounting and discusses perspectives on well-being in economics and psychology. Section 1.3 provides evidence on the link between self-reports of subjective well-being and objective outcomes, such as health and neurological activity. Section 1.4 introduces the evaluated time-use measures that we have developed and provides some evidence on their reliability and validity. Section 1.5 uses the PATS data to describe time use and affective experience across groups of individuals and activities. Section 1.6 provides a method for grouping activities into categories based on the emotional experiences that they are associated with. To illustrate the utility of our techniques, section 1.7 describes long-term historical trends in the desirability of time use and section 1.8 provides a cross-country comparison. Section 1.9 concludes by considering some knotty unresolved issues and by pointing to some opportunities for NTA in the future. 1.2 Conceptual Issues 1.2.1 Economics of Time Use, Goods, and Utility In a standard economic model, households receive utility from their consumption of leisure and goods. People choose to work because of the income and hence, consumption of goods that work makes possible. Available time and the wage rate are the constraints that people face. The national income and product accounts only value market output (or, equivalently, paid inputs and profits). Some attempts have been made to value nonmarket time using the wage rate as the shadow price of leisure. Becker (1965) argued that households combine resources (e.g., food) and time to produce output
14 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
(e.g., meals), just like firms. Thus, in Becker's model cooking only affects utility through the subsequent enjoyment of eating. Pollak and Wachter (1975) expand this framework to allow home production activities to affect utility through their direct effect on utility during the activities themselves and through the consumption of the output produced during the activities. Dow and Juster (1985) and Juster, Courant, and Dow (1985) emphasize the notion of "process benefits," or the flow of utility that accrues during particular activities, such as work and consumption.7 Juster, Courant, and Dow illustrate this idea in a Robinson Crusoe economy. Robinson can divide his time among three distinct activities: working in the market, cooking, and eating. He is constrained by the amount of food or clothing he can obtain through work, the amount of meals he can cook in a given period of time, and twenty-four hours in a day.8 With the assumption that process benefits from activities are separable, utility can be written as:
(1)
U Vm(tw,xc) Vc(tc,xc,xf) Ve(te,xc,xm),
where Vw, Vc, and Ve are the process benefits derived during work, cooking, and eating, respectively; xc is the quantity of clothing; xf is the quantity of food; xm is the amount of meals cooked; and t is the amount of time devoted to each activity. Juster, Courant, and Dow make the critical but sensible assumption "that the process benefit obtained from each activity is independent of the time and goods devoted to other activities" (128). They defend this assumption by noting that "any stocks produced by activity i are permitted to affect the process benefits from other activities."9 The data that we collect are divided into episodes of varying length, not activities, so it is more natural to model the time devoted to episodes and the average process benefit during those episodes. Consider someone who spends her first t1 hours of the day working, her next t2 hours preparing meals, her next t3 hours eating the meals prepared earlier, and her final t4 hours working again. (Of course, this could easily be extended to allow for more episodes and other activities.) Under the assumption of separability, the utility function can be written as:
1
2
3
4
(2) Ui v1(t,Xc)dt v2(t,Xc,Xf)dt v3(t,Xc,Xm)dt v4(t,Xc)dt.
0
0
0
0
Taking means of the flow utilities over the relevant intervals gives:
7. They define process benefits as the "direct subjective consequences from engaging in some activities to the exclusion of others. . . . For instance, how much an individual likes or dislikes the activity `painting one's house,' in conjunction with the amount of time one spends in painting the house, is an important determinant of well-being independent of how satisfied one feels about having a freshly painted house." The idea of process benefits is closely related to Kahneman's notion of "experienced utility." 8. We ignore sleep to simplify the exposition. 9. An exception might be exercise. A period of exercising may raise someone's mood during the rest of the day. We return to this following.
National Time Accounting: The Currency of Life 15 (3) Ui t1v1(t1,Xc) t2v2(t2,Xc,Xf) t3v3(t3,Xc,Xm) t4v4(t4,Xc). It follows that a person's total utility can be obtained from the duration weighted sum of average process benefits during the time the individual is engaged in each episode. There is no need to collect additional information on resources, constraints, or prices to summarize the person's well-being. Notice also that equation (3) does not require utility maximization. Even if the individual allocates his or her time suboptimally, if the mean process benefit can be estimated it is possible to estimate his or her well-being. In this framework, which loosely guides our empirical work, the average well-being among N members of society, W, is W Ui /N. If one wants to put a dollar value on W, in principle it is possible to estimate the monetary price that people are willing to pay on the margin to increase their process benefit in some activity by one unit, and use the inverse of this figure as a numeraire. For example, the way workers trade off pay for a more or less pleasant job can give an estimate of the marginal willingness to pay to improve time spent in a pleasant state. Alternatively, the amount that people are willing to spend on various types of vacations can be related to the flow of utility they receive during those vacations to place a monetary value on additional utility. Although it is possible, under the assumption of rational decision making, to place a dollar value on W in this framework, we shy away from this step and focus instead on providing credible estimates of W. Of course, measuring the flow of utility or emotions during various activities is no easy task, and some scholars doubt its feasibility entirely. Juster (1985) attempts to measure process benefits by using responses to the following question: "Now I'm going to read a list of certain activities that you may participate in. Think about a scale, from 10 to zero. If you enjoy doing an activity a great deal, rank it as a `10'; if you dislike doing it a great deal, rank it as a `0'; if you don't care about it one way or the other, rank it in the middle as `5'. . . . Keep in mind that we're interested in whether you like doing something, not whether you think it is important to do." The activities included: cleaning the house, cooking, doing repairs, taking care of your child(ren), your job, grocery shopping, and so forth. For activity j, the enjoyment score is assumed to equal the process benefit, Vj. There are several important limitations to Juster's type of enjoyment data, which we describe as a "general activity judgment" measure, because it focuses on a general response to a domain of life, not specific events that actually occurred. First, respondents are likely to develop a theory of how much they should enjoy an activity in order to construct an answer to the question. Second, respondents may be sensitive to the interviewers' reactions to their answers. For example, someone may be concerned that they will be viewed as a bad parent or worker if they respond that they do not like taking care of their children or their job. Third, people are unlikely to correctly aggregate their experiences over the many times that they engaged in
16 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone a particular activity in providing a general activity judgment. Other research (e.g., Kahneman, Wakker, and Sarin 1997) has found that individuals ignore the duration of events and instead place excessive weight on the end and peak of the experience when answering general evaluative recall questions. Fourth, and related, individuals are likely to exercise selection bias in choosing from the best or worst moments of past incidents of the specified activities. Results presented below cast some doubt on the validity of general activity judgments. Fifth, it is unclear if individuals utilize the enjoyment scales in an interpersonally comparable way. Nonetheless, as a description of time use and well-being, the process benefit approach has many advantages. Most importantly, the output of home production does not have to be observed or evaluated. A major goal of our work, therefore, has been to develop more informative measures of the flow of emotional experience during specific moments of the day. 1.2.2 The Psychology of Well-Being Contemporary psychology recognizes a variety of informative subjective well-being (SWB) measures. Our view of the structure of subjective wellbeing concentrates on two qualitatively distinct constituents that both contribute to SWB. The first component pertains to how people experience their lives moment to moment as reflected in the Positive and Negative feelings that accompany their daily activities. We refer to this component as "experienced happiness," or the average of a dimension of subjective experience reported in real time over an extended period. The second component pertains to how people evaluate their lives. It is typically assessed with measures of life-satisfaction, like "Taking all things together, how satisfied would you say you are with your life as a whole these days?" There are many ways in which these components of SWB can be measured, but we view them as reflecting overlapping but distinct aspects of people's lives. Much of the variance of both experienced happiness and life satisfaction is explained by variation in personal disposition that probably has a significant genetic component (Diener and Lucas 1999; Lykken 1999). We focus here on two other determinants: the general circumstances of people's lives (marital status, age, income) and the specifics of how they spend their time. Evaluating one's life as a whole poses a difficult judgment task (see Schwarz and Strack 1999). Like other hard judgments, the evaluation of one's life is accomplished by consulting heuristics--the answers to related questions that come more readily to mind (Kahneman 2003). Experimental demonstrations of priming and context effects provide evidence for the role of such heuristics in reports of life satisfaction (Schwarz and Strack 1999). Two heuristic questions that are used are: "How fortunate am I?" and "How good do I feel?" The first involves a comparison of the individual's circumstances to conventional or personal standards, while the second calls attention to
National Time Accounting: The Currency of Life 17 recent affective experience. Research indicates, for example, that reported life satisfaction is higher on sunny than on rainy days, consistent with the influence of the weather on their temporary moods. If individuals are first asked explicitly about the weather, however, they become aware that their current feelings may only reflect a temporary influence, which eliminates the effect of weather on reported life satisfaction (Schwarz and Clore 1983). In addition to personal effects, affective experience is determined by the immediate context and varies accordingly during the day; most people are happier sharing lunch with friends than driving alone in heavy traffic. Russell (1980) provides a theory of core affect, in which emotions are described along two dimensions. One dimension ranges from pleasure to displeasure, and the other from highly activated to deactivated. Happiness, for example, is an activated, pleasurable state. We define an individual's experienced happiness on a given day by the average value of this dimension of affective experience for that day. Experienced happiness, so defined, is influenced by the individual's allocation of time: a longer lunch and a shorter commute make for a better day. A person's use of time, in turn, reflects his or her circumstances and choices. Favorable life circumstances are more strongly correlated with activation than with experienced happiness. A classic puzzle in SWB research involves the limited long-term hedonic effects of outcomes that are greatly desired or feared in anticipation and evoke intense emotions when they occur (Brickman, Coates, and JanoffBulman 1978). In a recent study using longitudinal data, Oswald and Powdthavee (2005) find that average life satisfaction drops after the onset of a moderate disability but fully recovers to the predisability level after two years.10 This process is known as adaptation or habituation. Oswald and Powdthavee find that adaptation takes place but is incomplete for severe disabilities. Life events such as marriage and bereavement have substantial short-run effects on happiness and life satisfaction, but these effects are mainly temporary (e.g., Clark et al. 2003). Findings like these invite the idea of a potent process of hedonic adaptation that eventually returns people to a set point determined by their personality (see Diener, Lucas, and Scollon [2006]; Headey and Wearing [1989]). Kahneman and Krueger (2006) conclude that adaptation to both income and to marital status is at least as complete for measures of experienced happiness as for life satisfaction. This conclusion is also consistent with Riis et al. (2005), who used experience sampling methods to assess the feelings of end-stage renal dialysis patients and a matched comparison group. They found no significant differences in average mood throughout the day between the dialysis patients and the controls. 10. Smith et al. (2005) find that the onset of a new disability causes a greater drop in life satisfaction for those in the bottom half of the wealth distribution than for those in the top half, suggesting an important buffering effect of wealth, although low-wealth individuals still recovered some of their predisability well-being.
18 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone A focus on time use and activities suggests two factors in addition to hedonic adaptation for understanding the stability of SWB. First, although personality surely matters, the claim that an individual's experienced happiness must return to a set-point that is independent of local circumstances is probably false. For someone who enjoys socializing much more than commuting, a permanent reallocation of time from one of these activities to the other can be expected to have a permanent effect on happiness (Lyubomirsky, Sheldon, and Schkade 2005). Second, one must recognize that there are substantial substitution possibilities when it comes to activities. People who suffer injuries, for example, can substitute games like chess or checkers for competitive sports in their leisure time. These substitution possibilities are probably not anticipated. Thus, the largely unanticipated opportunity to substitute activities could attenuate the actual loss or gain in SWB associated with major changes in life circumstances, relative to anticipations. A final observation is that the withdrawal of attention is another mechanism of adaptation to life changes. Attention is normally associated with novelty. Thus, the newly disabled, lottery winner, or newlywed are almost continuously aware of their state. But as the new state loses its novelty it ceases to be the exclusive focus of attention, and other aspects of life again evoke their varying hedonic responses. Research indicates that paraplegics are in a fairly good mood more than half the time as soon as one month after their crippling accident. Intuitive affective forecasts will miss this process of attentional adaptation, unless they are corrected by specific personal knowledge (Ubel et al. 2005). 1.2.3 The U-Index: A Misery Index of Sorts Two challenges for developing a measure of the process benefit of an activity are that the scale of measurement is unclear, and different people are likely to interpret the same scale differently. Indeed, modern utility theory in economics dispenses with the concept of cardinal utility in favor of preference orderings. Survey researchers try to anchor response categories to words that have a common and clear meaning across respondents, but there is no guarantee that respondents use the scales comparably. Despite the apparent signal in subjective well-being data (documented in the next section), one could legitimately question whether one should give a cardinal interpretation to the numeric values attached to individuals' responses about their life satisfaction or emotional states because of the potential for personal use of scales. This risk is probably exacerbated when it comes to comparisons across countries and cultures. We propose an index, called the U-index (for "unpleasant" or "undesirable"), designed to address both challenges.11 The U-index measures the 11. The remainder of this section borrows heavily and unabashedly from Kahneman and Krueger (2006).
National Time Accounting: The Currency of Life 19
proportion of time an individual spends in an unpleasant state. The average U-index for a group of individuals can also be computed. This statistic has the virtue of being immediately understandable, and has other desirable properties as well. Most importantly, the U-index is an ordinal measure at the level of feelings. The first step in computing the U-index is to determine whether an episode is unpleasant or pleasant. There are many possible ways to classify an episode as unpleasant or pleasant. The data collected with Experience Sampling Methods (ESM) or the Day Reconstruction Method (DRM) include descriptions of an individual's emotional state during each episode in terms of intensity ratings on several dimensions of feelings, some of which are positive (e.g., "Happy," "Enjoy myself," "Friendly") and some of which are negative (e.g., "Depressed," "Angry," "Frustrated"). We classify an episode as unpleasant if the most intense feeling reported for that episode is a negative one--that is, if the maximum rating on any of the negative affect dimensions is strictly greater than the maximum of rating of the positive affect dimensions.12 Notice that this definition relies purely on an ordinal ranking of the feelings within each episode. Respondents can interpret the scales differently. It does not matter if respondent A uses the two to four portion of the zero to six intensity scale and Respondent B uses the full range. As long as they employ the same personal interpretation of the scale to report the intensity of their positive and negative emotions, the determination of which emotion was strongest is unaffected.13 It is reassuring to note that in cognitive testing conducted by the Bureau of Labor Statistics, ten subjects were asked whether the affective dimension that they gave the highest rating to was the most intense feeling they had during the episode, and all of the respondents said yes for each sampled episode.14 To define the U-index mathematically, let Iij be an indicator that equals 1 if a time interval denoted j of duration hij for person i is considered unpleasant and 0 otherwise. As mentioned previously, Iij equals 1 if the emotion that was rated as most intensive for that time interval is a negative one. For an individual, the U-index over a given period of time is jIijhij/jhij. For a group of N individuals, the U-index is defined as:
U i
j Iij hij j hij
/ N.
12. Our approach bears some resemblance to a procedure proposed by Diener, Sandvik, and Pavot (1991), which categorized moments as unpleasant if the average rating of positive emotions was less than the average rating of negative emotions. Unlike the U-index, however, averaging ratings of feelings requires a cardinal metric. Notice also that because the correlations between negative emotions tend to be low, their procedure will categorize fewer moments as unpleasant than the U-index. 13. Formally, let f( ) be any monotonically increasing function. If P is the maximum intensity of the positive emotions and N is the maximum intensity of the negative emotions, than f(P) f(N) regardless of the monotonic transformation. 14. Memo from Kathy Downey, research psychologist, Office of Survey Methods Research, BLS, July 21, 2008.
20 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Notice that the U-index for a group is the equally weighted U-index for the individuals in the group. The group U-index can be interpreted as the average proportion of time that members of the group spend in an unpleasant state. From a psychological perspective, the U-index has some desirable attributes. First, the predominant emotional state for the majority of people during most of the time is positive, so any episode when a negative feeling is the most intense emotion is a significant occurrence. It is not necessary to have more than one salient negative emotion for an episode to be unpleasant. Second, the selection of a negative feeling as more intense than all positive ones is likely to be a mindful and deliberate choice: the maximal rating is salient, especially when it is negative, because negative feelings are relatively rare. Third, because at a given moment of time, the correlation of the intensity among various positive emotions across episodes is higher than the correlation among negative emotions, one dominant negative emotion probably colors an entire episode and it is potentially misleading to average negative emotions. Of course, the dichotomous categorization of moments or episodes as unpleasant or pleasant obscures some information about the intensity of positive and negative emotions, just as a dichotomous definition of poverty misses the depths of material deprivation for those who are below the poverty line. However, we see the ordinal definition of unpleasant episodes as a significant advantage. In addition to reducing interpersonal differences in the use of scales, the question of how to numerically scale subjective responses is no longer an issue with our dichotomous measure. The categorization of moments into unpleasant and pleasant moments emphasizes what can be most confidently measured from subjective data. The U-index can be used to compare individuals (what proportion of the time is this person in an unpleasant emotional state?), demographic groups (do men or women spend a higher proportion of time in an emotional state considered unpleasant?), and situations. The U-index can also be aggregated to the country level (what proportion of time do people in France spend in an emotional state classified as unpleasant) and can be used to compare countries. Notice that because the U-index is aggregated based on time, it takes on useful cardinal properties. Like the poverty rate, for example, one could compute that the U-index is X percent lower for one group than another, or has fallen by Y percent from one year to another. 1.3 Is There Useful Signal in What People Report About Their Subjective Experiences? Economists often treat self-reported data with a high degree of suspicion, especially when those data pertain to subjective internal states, such as wellbeing or health. Is there any useful signal in what people tell us about their
National Time Accounting: The Currency of Life 21 subjective experiences? To answer this question, we first discuss how social scientists assess the validity of self-reports of behavior and subsequently develop a strategy for assessing the validity of self-reports of subjective experiences before we turn to relevant empirical findings. Following the review of the evidence, we identify some limiting conditions and highlight that self-reports of affect are most meaningful when they pertain to recent specific episodes in a person's life, a fact that we exploit later in the design of the Day Reconstruction Method and the Princeton Affect and Time-use Survey. 1.3.1 Rationale Many surveys ask respondents to report on their behavior. The validity of such reports can be assessed by comparing them with external records at the individual or aggregate level. For example, banking records can be used to evaluate the validity of self-reported expenditures at the individual level (e.g., Blair and Burton 1987), and national sales figures can be used to assess the validity of purchase reports in representative sample surveys at the aggregate level (e.g., Sudman and Wansink 2002). Neither of these strategies is feasible for assessing the validity of self-reported feelings, like moods, emotions, worries, or pain. Feelings are subjective experiences and the final arbiter is the person who experiences them. The same holds for other subjective evaluations, like reports of life-satisfaction, which pertain to individuals' subjective assessments of the quality of their lives. The subjective nature of feelings and evaluations precludes direct validation against objective records. It is also expected that comparisons of subjective and objective reports will not be identical, because people interpret the objective world in idiosyncratic ways. Nevertheless, one can gauge the validity of these reports in other, less direct ways. To begin with, one can assess interpersonal agreement: do "close others" perceive the person in ways that are compatible with the person's self-reports? While interpersonal agreement is comforting, it is less than compelling and subject to numerous biasing factors. As a more informative alternative, one can relate self-reports of subjective experience to objective outcomes with the expectation that there should be at least a modest correspondence. If reports of positive affect are associated with increased longevity, for example, they obviously capture something real--yet it remains unclear whether that something is indeed positive affect or some other variable correlated with its expression (the so-called "third variable" explanation). Perhaps people who present themselves in a positive light when answering questions also follow other strategies of social interaction that reduce daily friction and benefit health. Such ambiguities are attenuated when studies that do not rely on self-reports for the assessment of affect show similar results. Finally, interpretative ambiguities are further attenuated when experimental results, based on random assignment, support the
22 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone naturalistic observation; for example, when induced positive affect also has beneficial health consequences. Such supporting results will typically be more limited in scope due to ethical constraints on the experimental induction of affect (especially negative affective states such as stress or anger) and the more limited time frame of experimental studies. We next review illustrative findings from longitudinal studies that show self-reported affect predicts some important objective outcomes in life. Paralleling these naturalistic observations, a growing number of experimental studies documents compatible effects of induced affect, based on random assignment of participants to positive or negative "affect induction" conditions. For example, positive affect can be induced by giving subjects a cookie or placing a dime in a spot where they can find it. Other approaches to inducing affect include placing subjects in a situation where they overhear a compliment or insult, showing subjects a funny versus sad movie, asking subjects to recall a happy versus sad event, and giving subjects a task that is easy or impossible to perform; see Schwarz and Strack (1999). 1.3.2 Affect and Objective Outcomes: Social Life In a comprehensive review of cross-sectional and longitudinal studies, Lyubomirsky, King, and Diener (2005) observed that a preponderance of positive over negative affect predicts numerous beneficial outcomes, from the quality of one's social life and work life to longevity and the quality of one's health. Here, we focus on studies that are particularly informative with regard to the validity of affective self-reports, namely studies in which (a) the person's affect was assessed through self-reports several months or years prior to the observed outcome; (b) the outcome itself is objective (e.g., longevity or health status rather than subjective satisfaction with one's health); and (c) studies in which the affect assessment is not based on self-reports show compatible effects. Finding a Spouse Most people would prefer to be married to a partner who is happy and satisfied rather than depressed and dissatisfied. Consistent with this preference, several longitudinal studies show that people who report in sample surveys that they are happy (Marks and Fleming 1999) or satisfied with their lives (Lucas et al. 2003; Spanier and Fuerstenberg 1982) are indeed more likely to marry in the following years. For example, Marks and Fleming (1999) observed in a fifteen-year longitudinal study with a representative sample of young Australians that those who were 1 standard deviation above the mean of happiness reports were 1.5 times more likely to marry in the ensuing years; those 2 standard deviations above the mean were twice as likely to marry. This relationship can also be observed with measures of affect that do not rely on self-report. For example, Harker and Keltner (2001) coded the affect
National Time Accounting: The Currency of Life 23 expressed in women's college yearbook photographs, following the wellestablished procedures of Ekman's facial action coding system (Ekman and Rosenberg 1997). They observed that women who expressed genuine positive affect (in the form of a Duchenne smile) at age twenty-one were more likely to be married by age twenty-seven and less likely to remain single through middle adulthood. Of course, people may report being happy because they anticipate being married in the next year, but the long lag in the Ekman and Rosenberg study is harder to reconcile with reverse causality. Helping Others Several studies show that self-reported daily mood is associated with the likelihood of helping others. For example, Lucas (2001) observed that students who reported a preponderance of positive mood in their daily diaries also reported spending more time helping others than did those with less positive moods. Similarly, Csikszentmihalyi, Patton, and Lucas (1997) found that self-reported helping behavior increased with the percentage of time spent in a good mood among school-age youths. Numerous experimental studies, with random assignment to different affect induction conditions, support the link between positive mood and prosocial behavior. People in induced positive moods are more likely to help others by donating money (Cunningham, Steinberg, and Grev 1980), blood (O'Malley and Andrews 1983), and time (Berkowitz 1987) to worthy causes. Receiving a cookie or finding a dime is sufficient to elicit increased prosocial behavior (Isen and Levin 1972). Income Several studies show a positive relationship between self-reported positive affect at a given time and later income. Diener et al. (2002) observed that self-reported cheerfulness at college entry predicted income sixteen years later, controlling for numerous other variables, including parents' income. For example, the most cheerful offspring of well-off parents earned $25,000 more per year than the least cheerful offspring. Similarly, Marks and Fleming (1999) observed in their Australian panel study of young adults that respondents' self-reported happiness in one wave predicted the size of the pay raises they had received by the time of the next interview, two years later. Finally, Russian respondents who reported high happiness in 1995 enjoyed higher incomes in 2000 and were less likely to have experienced unemployment in the meantime (Graham, Eggers, and Sukhtankar 2006). 1.3.3 Affect and Objective Outcomes: Health Numerous longitudinal studies show that happy people have a better chance to live a long and healthy life (for reviews see Lyubomirsky, King, and Diener [2005]; Howell, Kern, and Lyubomirsky [2007]). This observation holds for mortality in general as well as for specific health outcomes;
24 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone moreover, it is supported by studies that relied on affect measures other than self-report. Mortality Based on data of the Berlin Aging Study, Maier and Smith (1999) reported that a preponderance of self-reported positive over negative affect (assessed with the Positive and negative affect schedule [PANAS]) predicted mortality in a sample of 513 older adults three to six years later. Studies with clinical samples reinforce this observation. For example, Devins et al. (1990) observed that end-stage renal patients who reported overall happiness were more likely to survive over a four year period than were their less happy peers. Similarly, Levy et al. (1988) found that women who reported more joy in life were more likely to survive a recurrence of Breast cancer over a seven year period. Studies based on personality tests that assess enduring affective predisposition replicate this conclusion (see Lyubomirsky, King, and Diener [2005] for a review). Complementary support for the observed relationship between positive affect and mortality comes from studies that asked the interviewer to rate the respondent's affective state. In one study (Zuckerman, Kasl, and Ostfeld 1984), healthy as well as unhealthy respondents who were rated as happier enjoyed lower mortality than their peers over a two-year period; Palmore (1969) replicated this observation over a more impressive period of fifteen years. Finally, in a study that attracted broad attention, Snowdon and his colleagues (Danner, Snowdon, and Friesen 2001; Snowdon 2001) analyzed autobiographical essays that young catholic nuns of the American School Sisters of Notre Dame had written in 1930, when most were in their early twenties. Coding the essays for emotional content, they discovered that positive affect expressed in these early essays was highly predictive of mortality by the time the writers were eighty to ninety years old. On average, nuns whose essays placed them in the top quartile of positive affect in the sample lived ten years longer than nuns whose essays placed them in the bottom quartile. Given that all nuns lived under highly comparable conditions in terms of daily routines, diet, and health care, this finding provides particularly compelling evidence for the repeatedly observed relationship between positive affect and longevity. Physiological Associations Several conceptual models in the fields of health psychology and behavioral medicine posit a central role for positive and negative affect in the translation of the psychosocial environment into physiological states and, subsequently, health outcomes, such as those mentioned previously. Empirical demonstrations of affect-physiology associations are a compelling source of validation for affect. We present representative findings in two physiological
National Time Accounting: The Currency of Life 25 systems--the immune system and the endocrine system--because of their close linkage with health outcomes. Immune Response Alterations in immune system functioning--either above or below normative levels--can result in greater susceptibility to invading organisms and neoplastic diseases, and to autoimmune conditions. Therefore, many studies have examined how psychosocial factors and affect are related to various compartments of the immune system. Several longitudinal studies observed that the frequency of self-reported hassles and uplifts and their accompanying affect is predictive of immune response. In one daily study, Evans et al. (1993) related participants' daily reports of life-events and mood over a two-week period to markers of immune function in daily saliva samples. They observed a higher secretion of immunoglobulin A on days that were characterized by many positive and few negative events. Stone and colleagues showed through their daily studies of events, mood, and symptoms that the impact of daily events on the secretory immune system was mediated through changes in negative and positive affect associated with daily events (Stone et al. 1987; Stone et al. 1996). A similar line of work by Vitaliano et al. (1998) monitored natural killer (NK) cell activity in cancer survivors. They found that participants who reported more uplifts than hassles (and presumably decreased levels of negative affect based on prior work [Stone 1987]) in daily life showed higher NK cell activity eighteen months later, an indicator of enhanced immune function. Moving to more major events, a classic extensive line of work by KiecoltGlaser and colleagues demonstrated that naturalistic situations such as students taking exams or maritally distressed individuals discussing their marital situation results in declines in immune functioning (e.g., KiecoltGlaser et al. 1988). Changes in the immune system have been shown by the same investigators to have health consequences, such as in the resolution of experimentally induced wounds. A particularly interesting series of studies by Cohen and colleagues demonstrated that people's level of affect is associated with their susceptibility to an experimentally induced viral infection and this is strongly supportive of the role of affect in physiology. In particular, recent evidence has indicated that proinflammatory cytokines are associated with positive affect (Doyle, Gentile, and Cohen 2006) when measured on a daily basis. Beneficial immune function effects of positive affect were also observed in experimental studies, based on random assignment to different affect induction conditions. For example, watching a humorous video clip has been found to increase NK cell activity and several other immune function markers (Berk et al. 2001), including salivary immunoglobulin A (Dillon,
26 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Minchoff, and Baker 1985) and salivary lysozyme (sLys) concentration (Perera et al. 1998). Induction of stressful situations has also produced changes in immune function. For example, Stone et al. (1993) exposed participants to challenging mental tasks and they subsequently had lower responsiveness of t-cells stimulated with standard antigens compared to participants who were not exposed. A recent review article by Marsland, Pressman, and Cohen (2007) concludes that positive affect is associated with up-regulation of the immune system. Hormones Many bodily functions are regulated by the actions of hormones, which are biological active substances secreted by various organ systems. One hormone that has been of particular interest to psychosocial researchers is cortisol, a product of the hypothamalic-pituatary-adrenal (HPA) system. Cortisol is often called the "stress hormone." It affects aspects of metabolism in general, but of special interest for this discussion is its impact of the immune system and its anti-inflammatory role. Observational and experimental studies have confirmed that cortisol levels are responsive to changes in affect and to experiences that are closely linked with affect changes. In an impressive line of research, Kirschbaum and colleagues (Kirschbaum, Pirke, and Hellhammer 1993) showed that a laboratory manipulation involving stressful student presentations quickly increased levels of cortisol; such changes could at least temporarily suppress the immune system. Supporting the experimental work, there is evidence from naturalistic studies that sampled respondents' affect and cortisol repeatedly throughout the day. Those studies showed that momentary negative affect is associated with higher levels of cortisol and positive affect with lower levels of cortisol (relative to when affect levels were at the opposite level) (Smyth et al. 1998). Furthermore, both state (momentary) and trait measurement of affect is associated in the same manner with cortisol levels (Polk et al. 2005). Neurological Activity Findings from neuroscience research also lend some support for the view that subjective reports are related to individuals' emotional states. By way of background, note that there is strong clinical and experimental evidence that the left prefrontal cortex of the brain is associated with the processing of approach and pleasure, whereas the corresponding area in the right hemisphere is active in the processing of avoidance and aversive stimuli. In particular, the left prefrontal cortex is more active when individuals are exposed to pleasant images or asked to think happy thoughts, while the right prefrontal cortex is more active when individuals are shown unpleasant pictures and asked to think sad thoughts. A study using several measures of
National Time Accounting: The Currency of Life 27 psychological well-being reported a statistically significant correlation of 0.30 between survey evidence on life satisfaction and the left-right difference in brain activation (Urry et al. 2004). In a striking demonstration of the validity of subjective reports, Coghill and colleagues compared subjects' self-reported pain levels to functional magnetic resonance imaging (fMRI) while applying a standardized pain stimulus to seventeen subjects. The pain stimulus consisted of hot presses against the lower leg. They found that individuals reporting higher levels of pain to the thermal pain stimulus produced greater activation of various cortical regions of the brain, some of which corresponded with the stimulated limb, than individuals who reported lower pain ratings to the same stimulus (see figure 1.1; Coghill, McHaffie, and Yen [2003]). The strong implication of this work is that variation in self-reports to standard stimuli are not simply a function of interpersonal differences in scale usage, but reflect, at least in part, differential neural processes associated with the perception of pain. They concluded, "By identifying objective neural correlates of subjective differences, these findings validate the utility of introspection and subjective reporting as a means of communicating a first-person experience" (8358). Other Systems Levels of positive and negative affect have also been associated with and shown to affect other physiological systems and we mention some of them here. Positive affect has been shown to increase performance on cognitive tasks and this could be associated with brain dopamine levels (Ashby, Isen, and Turken 1999). Relatedly, measures of brain activity have been associated with affective levels (Wheeler, Davidson, and Tomarken 1993). Some aspects of cardiovascular function and affect have been studied. Shapiro and colleagues (Shapiro, Jamner, and Goldstein 1997) used daily monitoring of affect and blood pressure to show that specific mood states such as anger were associated with increased levels of blood pressure. 1.3.4 Assessing Subjective Experiences As our review indicates, there is systematic signal in people's self-reports of their affective experiences. Nevertheless, self-reports of affect are subject to systematic methodological biases, which depend on the assessment method used. Next, we summarize what has been learned (for reviews see Robinson and Clore [2002]; Schwarz [2007]). When people report on their current feelings, the feelings themselves are accessible to introspection, allowing for more accurate reports on the basis of experiential information. But affective experiences are fleeting and not available to introspection once the feeling dissipated. Accordingly, the opportunity to assess emotion reports based on experiential information is limited to methods of momentary data capture (Stone et al. 2007) like
A. Frequency SI
ACC
B. z score SI
ACC
C. Frequency
D. z score
PFC
PFC
Frequency
z score
23456
3.89
14.5
Fig. 1.1 Brain regions displaying different frequencies of activation between highand low-(pain rating) sensitivity subgroups Source: Reproduced from: Coghill, McHaffie, and Yen (2003). Please see original image for references to color in the following note. Notes: Circles are centered on regions where the peak differences between groups were located. Colors in A and C correspond to the number of individuals displaying statistically significant activation at a given voxel (frequency), whereas colors in B and D correspond to the z-score of the subgroup analysis. Slice locations in A and B are ­2 mm from the midline, whereas slice locations in B and C are 32 mm from the midline (in standard stereotaxic space). Structural MRI data (gray) are averaged across all individuals involved in corresponding functional analysis.
National Time Accounting: The Currency of Life 29 experience sampling (Stone, Shiffman, and DeVries 1999), which we address in more detail in section 1.4. Once the feeling dissipated, the affective experiences need to be reconstructed on the basis of other information. When the report pertains to a specific recent episode, people can draw on episodic memory, retrieving specific moments and details of the recent past. Such reports can often recover the actual experience with some accuracy, as indicated by their convergence with concurrent reports (e.g., Kahneman et al. 2004; Stone et al. 2006). The Day Reconstruction Method, presented in section 1.4, takes advantage of this observation. In contrast, global reports of past feelings are based on semantic knowledge. When asked how they "usually" feel during a particular activity, people draw on their general beliefs about the activity and its attributes to arrive at a report. The actual experience does not figure prominently in these global reports because the experience itself is no longer accessible to introspection and episodic reconstruction is not used to answer a global question. Finally, the same semantic knowledge serves as a basis for predicting future feelings, for which episodic information is not available to begin with (Schwarz, Kahneman, and Xu 2009; Xu and Schwarz 2009). These hedonic predictions, in turn, often serve as a basis for behavioral choice (March 1978). These processes result in a systematic pattern of convergences and divergences in affect reports. First, concurrent reports and retrospective reports pertaining to specific recent episodes usually show good convergence, provided that the episode is sufficiently recent to allow detailed reinstantiation in episodic memory. Second, retrospective global reports of past feelings and predictions of future feelings also show good convergence, given that both are based on the same semantic inputs. Hence, global memories are likely to "confirm" predictions. Third, choices are based on predicted hedonic consequences, and are therefore usually consistent with predictions and global memories. However, fourth, global retrospective reports as well as predictions and choices will often diverge from concurrent and episodic reports, given that the different types of reports are based on different inputs. As a result, a person's expectations and global memories go hand in hand, but often fail to reflect what the person actually experienced moment to moment (for a review see Schwarz, Kahneman, and Xu 2009). These observations have important implications for the assessment of affective experience in time-use studies. They highlight that global reports of how much one usually enjoys a given activity are a fallible indicator of people's actual affective experience in situ. Such global reports were used in Juster and colleagues' pioneering studies (e.g., Juster and Stafford 1985). Our work builds on Juster's (1985) conceptual approach while heeding the lessons learned from recent psychological research by employing measures of affective experience that pertain to specific episodes of the preceding day. Next, we turn to the development of these measures.
30 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone 1.4 Methods for Collecting Evaluated Time-Use Data: From EMA to DRM to PATS The Experience Sampling Method (ESM) and Ecological Momentary Assessment (EMA) were developed to collect information on people's reported feelings in real time in natural settings during selected moments of the day (Csikszentmihalyi 1990; Stone and Shiffman 1994). Participants in real-time studies carry a handheld computer that prompts them several times during the course of the day (or days) to answer a set of questions immediately.15 Participants are typically shown several menus, on which they indicate their physical location, the activities in which they were engaged just before they were prompted, and the people with whom they were interacting. They also report their current subjective experience by indicating the extent to which they feel the presence or absence of various feelings, such as angry, happy, tired, and impatient. Momentary real-time surveys are often viewed as the gold standard for collecting data on affective experience because it minimizes effects of judgment and of memory. As a convention, we will refer to studies that collect data on emotions in real time as ESM studies throughout the remainder of the chapter (because we are focusing on experience rather than environmental features). So far, however, real-time data collection has proved prohibitively expensive and burdensome to administer to large, representative samples. An alternative to ESM that relies on a short recall period is the Day Reconstruction Method (DRM), which is described in Kahneman et al. (2004). The DRM combines elements of experience sampling and time diaries, and is designed specifically to facilitate accurate emotional recall.16 Respondents--who participated in the survey in a central location--were provided with four packets containing separate questionnaires, and were asked to answer them in sequence. The first packet had standard questions on life, health, and work satisfaction and demographics. Satisfaction questions were asked first so that answers were not contaminated by the other questions and diary that followed. Second, respondents filled out a time diary summarizing episodes that occurred in the preceding day. The third packet asked respondents to describe each episode of the day by indicating the following: when the episode began and ended, what they were doing (by selecting activities from a provided list), where they were, and with whom they were interacting. To ascertain how they felt during each episode in regards to selected affective dimensions, respondents were also asked to report the intensity of their feelings along twelve categories on a scale from zero ("Not at all") to six ("Very Much"). The affective categories were specified by descriptors, 15. Other survey technologies can also be used for EMA, such as paper diaries and cell phones. 16. Robinson and Godbey (1997), Gershuny and Halpin (1996), and Michelson (2005) have used data collected from related survey techniques.
National Time Accounting: The Currency of Life 31 Survey Techniques for Collecting Data on Evaluated Time Use Experience Sample Method (ESM) and Ecological Momentary Assessment (EMA). ESM and EMA are techniques for collecting data on time use and emotional experiences in real time. Respondents typically carry a computer device (a Personal Digital Assistant, called a PDA, for example) and indicate features of their activity and the feelings prior to being signaled by the device. EMA studies typically collect environmental information as well and may include physiological measurements (e.g., blood pressure, cortisol). Day Reconstruction Method (DRM). DRM is a paper-and-pencil questionnaire that first collects time diary information from individuals for the preceding day. The diaries can list personal details, as they are not collected. Then, for each indicated episode, individuals indicate the nature of the activity, who was present, and the extent to which various emotions were present or absent. Princeton Affect and Time Survey (PATS). PATS is a telephone survey patterned after the American Time Use Survey. After individuals report the activities of the preceding day (who with, what doing, where, when started and ended), three fifteen-minute intervals are randomly sampled and respondents are asked the extent to which various emotions were present or absent during that time. mostly adjectives, such as happy, worried/anxious, and angry/hostile. The anchor, "Not at all," is intended to be a natural zero point that has a common meaning across respondents for these descriptors. The final packet contained personality and work questions. Subjects were paid $75 for filling out the DRM questionnaire, which usually took forty-five to seventy-five minutes to complete. The emotions that respondents were asked to rate for each episode in the DRM were selected in part to represent points along the Russell (1980) affect circumplex. This distinguishes the DRM from the small number of past diary studies that included a question on how much individuals enjoyed (or liked/disliked) the activity they were doing. Russell models emotions as consisting of two core dimensions, pleasantness (pleasant versus unpleasant) and activation (aroused versus unexcited), with emotions positioned on a circle in this space. We interpret the duration-weighted average of the reported affect intensities as the average flow of "process benefits" or experienced well-being during the interval. An early version of the Day Reconstruction Method was applied to a sample of 909 working women in Dallas and Austin, which we refer to
32 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone as the Texas DRM (Kahneman et al. 2004).17 Another DRM survey was conducted of 810 women in Columbus, Ohio and 820 women in Rennes, France in the spring of 2005.18 A major goal of the Texas DRM study was to determine whether, despite its reliance on memory, the DRM reproduces results found in ESM. We looked in particular for features of experience captured by ESM and DRM that deviate from people's lay intuitions. If DRM reproduces these patterns we can conclude that it captures respondents' actual experiences during the preceding day rather than their general intuitions about what their experiences "must have been like." One comparison along these lines is shown in figure 1.2, which shows hourly mean ratings of "tired" in the DRM and from an independent study that used experience sampling. Whereas people's intuitions might hold that tiredness rises monotonically throughout the day, ESM studies show that tiredness reaches a minimum around noon. The DRM data replicate this V-shaped pattern, and the results obtained with ESM and DRM methods are remarkably similar. Moreover, this V-shaped pattern of tiredness was found in four subsequent DRM studies. Other results of the Texas DRM conformed reasonably well to basic results frequently observed in Experience Sampling, despite differences in the sample demographics.19 For example, the incidence of negative emotions is relatively rare in DRM--"angry/hostile" was rated above zero only 23 percent of the time, while feeling "happy" was rated above zero 95 percent of the time. The same pattern is found in ESM studies. The correlations among the emotions, particularly the positive ones, were quite high across episodes--around 0.7 for positive emotions and 0.4 for negative emotions. This pattern also replicates ESM findings. For example, the correlation of happy and "enjoying myself " across episodes is 0.73 in the DRM and 0.80 for a specialized sample of arthritis patients who participated in an ESM study.20 Unfortunately, we are not aware of a real-time data capture study that collected suffficiently comparable data to compare activity ratings in the two methods. Though not definitive, these findings suggest that DRM provides a reasonable approximation to the results of the more demanding ESM. We also compared the DRM to a set of general activity judgment questions that closely replicated Juster (1985). Specifically, we asked the following questions shown in table 1.1 to 252 women in Texas in 2002 who were recruited in the same fashion as the Texas DRM sample. 17. The sample consisted of 535 respondents who were recruited through random selection from the driver's license list plus a screen for employment and age eighteen to sixty, and another 374 workers in three occupations: nurses, telemarketers, and teachers. Because most results were similar for both subsamples, we present results for the full sample. 18. Sampled individuals were identified by random-digit dialing. 19. See Kahneman et al. (2004) for further examples of nonintuitive patterns obtained with both methods. 20. This correlation was computed using a sample of eighty-four arthritis patients who were prompted to report their feelings on a zero to 100 visual analog scale three to twelve times a day, over an entire week.
National Time Accounting: The Currency of Life 33 Fig. 1.2 Comparison of pattern of tiredness over the day based on DRM and ESM samples Source: Kahneman et al. (2004). Note: Points are standard scores computed across hourly averages within each sample. We then used just the adjective "enjoy" on a zero to six scale from the Texas DRM to compute the average reported enjoyment while women engaged in these thirteen activities according to the diary study. Table 1.2 compares the ranking of activities from the two approaches. The correlation between the ranks is 0.69. With small samples and some possible differential selection as to who participated in the activities on the diary day, the results should be read cautiously. Still, the results of the global ratings are quite similar to Juster (1985). The original Juster survey found that work and child care ranked particularly highly in terms of enjoyment, while our replication survey finds a similar result, especially for child care. More important, however, the DRM affect reports paint a different picture. For example, child care is reported as more enjoyable when asked about as an activity than in the diary-based study.21 Work is ranked eighth in the Juster-like survey, 21. Robinson and Godbey (1997) found a similar result comparing his diary-based study to Juster's ranking.
34 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.1
Juster-like question in our replication survey
We would like to learn how likable or dislikable various activities are. Below we list a number of different things that you may often likely to do in your life. For each one, please circle the response that indicates how much you like or dislike it: (if one does not apply to you, you may skip it)
Dislike a great deal
Like a great deal
Commuting to work
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Working in your main job
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Having lunch on a workday
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Socializing at work
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Commuting to home from work ­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Socializing with friends
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Talking on the phone at home ­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Taking care of your children
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Doing housework
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Cooking/preparing food
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Having dinner on a workday
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Relaxing at home
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
Watching TV
­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5
perhaps not as highly as in the original because of our focus on women, but still higher than in the DRM. Interestingly, socializing after work is ranked much more highly in the DRM than in the general activity question. The contrast between these results, together with the contrast between the DRM and the original Juster rankings of activities, highlights the importance of collecting event-based data. Asking people to respond about how they feel about activities in general tends to provide a different ranking than when their actual experiences are used to guide their reported feelings during those activities (for a more detailed discussion see Schwarz Kahneman, and Xu 2009).22 1.4.1 PATS: A Phone Survey Version of DRM The DRM is also burdensome and difficult to implement in a national sample. We designed the Princeton Affect and Time Survey to collect data 22. Gershuny and Halpin (1996) also cast doubt on the utility of general activity judgments. They analyzed data from a survey of British married couples in 1986 that asked a set of general questions about enjoyment with various activities. Respondents also maintained a diary for five days in which they reported their main activity during thirty-minute intervals and, for each interval, how much they enjoyed their main activity, on a scale of 1 (very much) to 5 (not at all). Looking across subjects for a given activity, the proportion of the variation in the diaryderived enjoyment scale explained by the corresponding general activity enjoyment response was low, only 11 percent for supervising kids and 10 percent for cooking. Thus, people did a poor job predicting their own reported emotional experiences with a general activity enjoyment question.
National Time Accounting: The Currency of Life 35
Table 1.2
Rank of activities in terms of average enjoyment from DRM and general activity enjoyment question similar to Juster (1985)
Activity
DRM (enjoy)
Juster enjoy/dislike
Child care
9
2
Commuting from work
12
11
Commuting to work
13
13
Cooking
8
9
Dinner
3
3
Housework
10
12
Lunch
4
4
Phone at home
7
10
Relaxing
2
1
Socializing after work
1
7
Socializing at work
6
5
Watching TV
5
6
Working
11
8
from respondents over the phone more expeditiously. A related goal was to develop a module that could be added to the U.S.' main time-use survey, the ATUS. The PATS survey works as follows. We started with the BLS ATUS questionnaire and eliminated a small number of questions that were not relevant. Respondents were first asked to describe each episode (defined as an interval of time in which the respondent was engaged in a specified activity; the average respondent reported 17.8 episodes) of the preceding day, using the ATUS protocols. Information about the activity individuals engaged in--what they were doing, where they were, and who was with them--was collected for each episode. After the entire day was described in this manner, three episodes were randomly selected in proportion to duration and without replacement.23 For these episodes, respondents were asked a five-minute module of questions, covering the extent to which they experienced six different feelings (pain, happy, tired, stressed, sad, and interested) during each episode on a scale from zero to six. They were instructed that a zero meant they did not experience the feeling at all at the time and a six meant the feeling was very strong. Specifically, respondents were asked to report their feelings during a randomly selected fifteen-minute interval of the sampled episodes. They were also reminded of what activity they said they were doing at that time in the diary part of the questionnaire. The order in which the feelings were presented was randomly assigned across respondents from six different permutations. The sampled episodes were ordered chronologically in the 23. More specifically, the BLAISE computer program divided the day into fifteen-minute intervals and randomly selected three fifteen-minute intervals. If any of those intervals was in the same episode, additional fifteen-minute intervals were selected that were in other episodes so an episode was only included at most once.
36 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone module. We also collected information on whether the individual was interacting with someone during sampled episodes. The adjectives used in the PATS only partially overlap with those used in our DRM studies for a few reasons. First, we asked a smaller number of adjectives to save respondent time. Second, we avoided using compound adjectives, which we thought could be confusing to respondents over the phone. Third, the Gallup Organization conducted a set of twenty-five cognitive interviews with respondents to check their understanding of the affect questions and to make sure the questions made sense during most nonsleeping activities. These interviews helped us narrow down the set of emotions asked about. The survey was administered by the Gallup Organization on our behalf in a random digit dial telephone survey of U.S. residents from May to August of 2006. Interviews were conducted in English and Spanish. A total of 3,982 people completed the survey, for a response rate of 37 percent. Weights were developed by Gallup to make the sample representative of the general population in terms of geographic region, gender, age, and race. The weights were based on counts from the Current Population Survey (CPS). Sixty-one percent of the unweighted respondents were women, a majority were white (88 percent), 90 percent had a high school education or higher, and 40 percent had household income less than $40,000 per year. The average age was 51.4 years. Reweighting the sample to represent the population resulted in some significant distributional changes. Most notably, compared with the unweighted sample, the weighted sample had fewer women (53 percent), higher income (36 percent below $40,000), and a lower average age (45.2 years). Unless otherwise noted, we apply sample weights in all of the statistics we report based on PATS. 1.4.2 Evaluating PATS We will use the PATS to illustrate NTA, so it is important to evaluate its properties in comparison to other time-use data sets and in comparison to results for affective experience captured in ESM and DRM. Figure 1.3 shows that the allocation of time across activities (weighting individuals by sample weights) from the PATS closely matches that in the ATUS for the same months of 2004 and 2005. The correlation between time spent in these activities from the two surveys is an impressive 0.99. This high concordance suggests that the weighted sample is representative of the population, at least in terms of time use. In figure 1.4 we show the distribution of responses to the questions about feeling happy and tired over episodes in the PATS and Texas DRM. These adjectives were selected because they display different patterns--strongly skewed to the left for happy and slightly skewed to the left for tired except for a prominent mode at zero. It is reassuring that the distributions are very similar in both methods. Moreover, the incidence of reports of negative emotions was rare in PATS as was found in DRM and ESM.
National Time Accounting: The Currency of Life 37 Fig. 1.3 Average hours per major activity in PATS and ATUS Notes: PATS shown in black and ATUS in white. PATS was conducted in May­August 2006 and ATUS is for May­August 2004­05. Fig. 1.4 Distribution of reported happiness and tired in PATS and DRM We can also compare correlations between feelings across episodes in PATS to those in DRM and ESM. The correlation between feeling happy and feeling tired, for example, is ­0.13 for women in the PATS, ­0.21 in the Texas DRM survey of women, and ­0.34 in a Columbus, Ohio DRM survey of women. The correlation between feeling happy and stressed is ­0.29 across women's episodes in PATS, and ­0.44 in the Columbus DRM.
38 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone The correlation between pain and happiness across episodes in the PATS is ­0.10, while the corresponding correlation across moments in ESM data is ­0.20 for the sample of arthritis patients mentioned previously. These results suggest that the correlation between pairs of reported emotions in the PATS is a little weaker than the corresponding correlations in ESM and DRM, but they point in the same direction and are qualitatively similar. With only three sampled episodes per interview, it is probably more difficult for respondents to reproduce their precise pattern of tiredness over the day. Still, the correspondence between the diurnal pattern of tiredness in PATS and DRM and ESM is reasonable (see fig. 1.5). The pattern displayed by the PATS data is much less V-shaped than was the case in the other surveys, but the increasing pattern of tiredness in the afternoon and evening is clearly evident. The correlation between the average rating of tiredness each hour in PATS and DRM is 0.87, and between PATS and ESM is 0.86. Moreover, the PATS data show similar age interactions to what we found earlier; namely, a sharper decline in tiredness in the morning for younger respondents. The correlation between reported life satisfaction and net affect across people was also similar in PATS and the Texas DRM. In the (random sample component of the) Texas DRM, the correlation between life satisfaction and net affect is 0.44 and in the PATS it is 0.35. Because net affect can be computed for only three episodes per person in the PATS, however, one would expect the 0.35 correlation to be biased downward. To make a fairer comparison, we randomly selected three episodes per person from the DRM. In this more comparable sample, the correlation fell to 0.39, quite close to the 0.35 computed with PATS. Krueger and Schkade (2008) provide estimates of the reliability of life satisfaction and net affect. Using their estimates to adjust for attenuation bias, the correlation between life satisfaction and net affect would rise from 0.44 to around 0.70. This figure suggests that interpersonal variations in average net affect over many days reflects about half of the variability in life satisfaction. Table 1.3 considers how the average rating of happy compares across common activities in the PATS and the random sample of the Texas DRM, both on a zero to six scale.24 The Pearson correlation between the two measures is 0.78, and the rank-order correlation is 0.74. Childcare is the largest outlier, with a one-half point lower rating in the DRM. Television is another outlier, with the DRM exceeding the PATS.25 In these respects, the PATS ranking of activities are intermediate between the rankings in the Juster-like survey and the DRM. It is possible that in the PATS, respondents reflect more on the activity in general than the particular episode. Another possibility is that 24. Attempts were made to make the activities as comparable as possible. 25. See Kubey and Csikszentmihalyi (1990) for a real-time study of subjects' emotional experiences while watching television.
Fig. 1.5 Comparison of pattern of tiredness over the day based on PATS, DRM, and ESM samples Notes: Points are standard scores computed across hourly averages within each sample.
Table 1.3
Comparison of PATS and DRM average happiness rating (0­6) by activity
Activity
PATS
DRM
Difference
Housework
3.77
4.10
Commuting
3.80
3.84
Working
3.82
3.74
Watching TV
3.91
4.32
Computer
4.06
3.94
Shopping
4.11
4.00
Preparing food
4.25
4.27
On the phone
4.47
4.00
Relaxing
4.49
4.55
Eating
4.57
4.43
Child care
4.59
4.06
Socializing
4.74
4.48
Prayer/worship
4.97
4.56
Exercising
5.09
4.77
Unweighted average
4.37
4.23
­0.33 ­0.04 0.08 ­0.41 0.12 0.11 ­0.02 0.47 ­0.06 0.14 0.53 0.26 0.41 0.32 0.15
Notes: PATS sample is men and women combined. DRM sample is random component of Texas survey.
40 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.4
Average response by order of affect questions in PATS sample
Average
Happy
Tired
Stressed
Sad
Interested
Pain
Question order
First
4.35
2.31
1.37
0.71
4.34
0.89
Second
4.22
2.62
1.41
0.68
4.10
0.97
Third
4.19
2.67
1.62
0.69
3.90
0.98
Fourth
4.18
2.65
1.58
0.83
3.92
0.96
Fifth
3.88
2.67
1.49
0.70
4.10
1.03
Sixth
3.99
2.71
1.54
0.69
4.07
1.08
All
4.13
2.61
1.50
0.72
4.07
0.99
Notes: One of the following six different orderings was randomly selected for each respondent. Order 1: Happy, Tired, Stressed, Sad, Interested, Pain; Order 2: Tired, Stressed, Sad, Interested, Pain, and Happy, Order 3: Stressed, Sad, Interested, Pain, Happy, and Tired; Order 4: Sad, Interested, Pain, Happy, Tired, and Stressed; Order 5: Interested, Pain, Happy, Tired, Stressed, and Sad; Order 6: Pain, Happy, Tired, Stressed, Sad, and Interested. Results are unweighted.
differences in the sample populations between PATS and the DRM account for the discrepancies. Table 1.4 summarizes results on how the order of emotions affected reported intensity of feelings in PATS. As mentioned, we randomly assigned respondents to one of six different orderings for the affect questions. Once an order was selected, the same order was used for each of the three sampled fifteen-minute intervals. The order effect for each of the emotions is statistically significant at the 0.025 level, and usually much lower. As a general rule, when positive emotions were asked about early on, their ratings tended to be higher, and when negative emotions were asked about early on, their ratings tended to be lower. If happy was asked first, for example, its mean response was 4.35, compared with 3.99 when it was asked last; when pain was asked first its mean response was 0.89, compared with 1.08 when it was asked last. Interestingly, the adjective "interested" behaved like a positive emotion in this regard. Table 1.2 combines results for the first, second, and third episode that was inquired about. Surprisingly, when we disaggregated the order effects were not notably stronger for the first of the three episodes. We expected to find stronger order effects for the first episode, as the order was known to respondents by the second and third episode. One interpretation of these results is that the first emotion provides an anchor for the subsequent ones. Respondents are typically in a positive mood before the affect questions are asked (judging from the high frequency of positive affect), and the response to the first emotion question is anchored relative to this positive feeling. Because the order in which emotions were presented was randomly assigned to respondents in PATS, our results should not be biased by order effects in any event.
National Time Accounting: The Currency of Life 41 It is also worth noting that the particular ordering used did not have a significant effect on the level of the U-index ( p-value 0.37 for joint F-test of constant U-index). Thus, a salutary feature of the U-index is that it is apparently robust to order effects, because the anchoring that produces the order effects does not substantially alter the ordinal ranking of emotional ratings. We can examine how the weather relates to the PATS affect and satisfaction data. Table 1.5 summarizes results from Connolly (2007), who merged daily weather data from the National Climate Data Center to the PATS survey. Specifically, she merged data on the mean temperature and amount of rainfall on the interview day and diary day (which is the day prior to the interview day), as well as the normal temperature and rainfall for the season and geographic area. Because temperature is highly correlated on adjacent days, it was not possible to estimate separate effects of the temperature on the interview and diary day. Rainfall, however, varies considerably from day to day. Women's reports of their life satisfaction and affect were more sensitive to the weather than men's, so we focus on results for women here. As in Schwarz and Clore's (1983) survey, Connolly found that life satisfaction was lower in the PATS if women were interviewed on rainy days. Life satisfaction was also lower in areas with higher normal precipitation levels and temperature. Temperature on the interview day was unrelated to life satisfaction, but a higher temperature on the diary day was associated with lower net affect. Since PATS was conducted in the late spring and summer, one might expect hotter days to be associated with lower net affect. Rain on the interview day was insignificantly related to net affect, while a small amount of rain on the diary day was associated with lower net affect. These
Table 1.5
Summary of effects of weather on reported well-being in the PATS survey
Variable
Life satisfaction
Net affect
Normal rainfall
--
0
Rain on interview day
--
0
Rain on diary day
0
­/0
Normal temperature

0
Temperature on interview day
0
n.a.
Temperature on diary day
n.a.
--
Notes: Connolly entered dummy variables for ranges of the rain and temperature variables in her regression analysis. A negative sign here indicates a negative and statistically significant effect of the climate measure, a positive sign indicates a positive and statistically significant effect of the climate measure, and n.a. indicates that the measure was not included in the particular analysis because of multicolinearity. Sample consists of women from PATS. The satisfaction regression also controlled for demographic variables (education, age, marital status, race, and ethnicity). The net affect regression also controlled for activity dummies, month, day, state, and demographic variables. See tables 3.4, 3.12, and 3.16 of Connolly (2007) for the underlying estimates.
42 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone results suggest that the weather influences reported net affect in the PATS data in a plausible way that is consistent with the true effect of the weather on people's moods, while the weather on the interview day is unrelated to net affect reported for the preceding day, as one would hope. Finally, Alan Krueger and Arthur Stone have conducted a small scale study of 168 workers in Syracuse, NY and Stony Brook, NY who participated in a specially designed ESM study on three consecutive days in the spring and summer of 2008 (on a Thursday, Friday, and Saturday). A day later, participants also completed the PATS questionnaire referring to the same days covered by the ESM survey. In the ESM component of the survey, respondents were asked about their feelings on six occasions on each day, after being prompted by a PDA. The PATS component asked about emotions during three randomly selected fifteen-minute intervals. Because it proved impossible to conduct the study on a representative sample, subjects were recruited through advertising and were offered $120 for their participation. But because we compare reported emotions from the two survey modes for the same individuals, any systematic differences are likely to be due to the survey methods. To avoid confusion, we call the PATS component of this survey PATS-2. The PATS-2 interviewing was also conducted by Gallup. The emotions inquired about in the PATS-2 and ESM questionnaires included those in the original PATS (happy, sad, stressed, pain, etc.). We use these data to compare the real-time responses of respondents to their recalled experiences in the PATS-2 instrument. Figure 1.6 reports the average rating of the emotions from the two surveys. The negative emotions received a slightly higher rating in the ESM than in the PATS-2 survey, which may partly reflect their order on the ESM questionnaire (in the PATS-2 the order was randomly assigned). The differences Fig. 1.6 Average of subjects' ratings in ESM and PATS-2 for same sample members Notes: Order of emotions was randomized in PATS-2. Sample is 165 individuals who responded in both surveys. Except for happy, all differences are significant at 0.005 level in paired t-test.
National Time Accounting: The Currency of Life 43
are qualitatively small, however, even though they are usually statistically significant. Clearly, the pattern of intensity across emotions is the same regardless of whether the emotions were recalled or collected in real time. For the 105 moments in time that were sampled in both the ESM and PATS-2 surveys (those that by chance happen to have overlapped), we can calculate the correlation between the emotions from the two surveys. The correlations ranged from 0.41 for happiness to 0.54 for pain. The correlation of the U-index measured in overlapping moments was 0.54. These correlations are lower than one might hope, but still nontrivial. Moreover, they could be biased slightly downward because the PATS refers to a fifteenminute slice while the ESM data are for a moment in time. A larger sample can be used to compare the ratings of activities because it is not necessary to restrict the sample to overlapping moments. Table 1.6 reports the U-index during various activities for the two survey modes. We restrict the sample to activities with at least forty-five sampled episodes in PATS-2 to reduce sampling error. In both survey modes the U-index is low for social activities and eating, and high for work and travel time. The correlation of the measures across the activities is 0.83, and the rank correlation is 0.86. Given the sampling variability inherent in the activitylevel U-indices, it is also noteworthy that if we weight the activities by the PATS-2 sample size (which ranges from forty-five to 423), the correlation rises to 0.90. Finally, we note that we used the ESM-PATS-2 data to compute the correlation of person-level averages. That is, for each individual we computed the average of the (up to eighteen) ESM ratings and of the (up to nine) PATS-2 ratings of each emotion, and computed the correlation between them. The correlation ranged from 0.75 for happiness to 0.86 for pain. These correlations are attenuated by sampling variability, however, as we only sampled a small number of random moments from each person's day. If the correlation is adjusted for sampling variability, it rises to 0.92 for happiness and 0.94 for pain.
Table 1.6
Average U-Index during popular activities, as measured by ESM and PATS-2 for the same sample
Activity
ESM
PATS-2
Work Housework Socializing Travel Grooming Eating/drinking Recreation All activities
0.157 0.093 0.088 0.143 0.156 0.080 0.114 0.126
0.156 0.117 0.076 0.144 0.133 0.043 0.068 0.112
Notes: U-index equals one if rating of stress, sad, or pain exceeds happiness. Activities are based on PATS-2 questionnaire.
44 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone We conclude that the PATS instrument and real-time reporting do a reasonably similar job characterizing individuals or activities. They are less consistent in describing feelings at specific moments, although the measures are still positively correlated and the mean reported emotion over all moments is remarkably similar regardless of whether it is reported in real-time or recalled a day later. 1.5 Well-Being across Groups and Activities 1.5.1 Differences in Well-Being between Groups We use the PATS to compare affective experience across groups of individuals and frequent uses of time. Table 1.7 reports the average U-index for several demographic groups, and some of those results are highlighted here. (Table 1A.1 presents results of the effect of demographic and other variables on the U-index in a multiple regression framework.) The U-index is 2 points lower for men than women ( p-value 0.10). The U-index is higher for blacks and hispanics than for whites. The U-index falls with household income and education. Those in households with income below $30,000 per year spend almost 50 percent more time in an unpleasant state than do people with income above $100,000 per year (22.5 percent versus 15.7 percent). The data indicate a mild inverse U-shape pattern in unpleasant moments with age for women. These patterns are often found in life satisfaction data and in our earlier DRM studies. Married men and women have the same U-index, 17.4 percent. The U-index for never married men and cohabiting men is also around 17 percent. The U-index is notably higher for unmarried women and divorced men. The former result is a contrast to our previous DRM studies, which found that married and unmarried women exhibited a similar U-index. Interestingly, the U-index is around 23 percent for all groups of unmarried women, divorced, widowed, cohabiting, and never married. In a regression, the married-unmarried gap is not accounted for by controlling for demographic variables or activities. Controlling for differences in household income, however, accounts for more than half of the marriage gap in the U-index for women. 1.5.2 Activities Table 1.8 reports the U-index and mean of five reported emotions during various primary activities. The order of activities is ranked by the U-index. The U-index is relatively low during discretionary activities, including religion/prayer, sports and exercise, relaxing and leisure, and socializing. Watching television is rated in the middle of the activities shown, as are food preparation and volunteering. The highest U-index activities include housework, working for pay, household management, receiving medical care, edu-
National Time Accounting: The Currency of Life 45
Table 1.7
U-Index for various demographic groups, PATS data
Demographic
(%)
Sex
Men
17.6
Women
19.6
Race/ethnicity
White
17.5
Black
23.8
Hispanic
21.9
Household income
$30,000
22.5
$30,000­$50,000
18.6
$50,000­$100,000
18.6
$100,000
15.7
Education
High school
20.5
High school
21.3
Some college
19.6
College
15.6
Master's
16.6
Doctorate
11.3
Men (%)
Women (%)
Age
15­24
18.8
18.9
25­44
17.1
20.5
45­64
18.7
20.9
65
15.6
16.1
Marital Status
Married
17.4
17.4
Divorced/separated
24.3
24.5
Widowed
20.2
22.3
Never married
16.9
23.2
Cohabiting
17.3
23.3
Notes: U-index is proportion of time that rating of sad, stressed, or pain exceeds happy.
cation, and caring for adults. This pattern is quite plausible, although it deviates in some important respects from the Juster-like general activity results. Some of the ratings of the specific emotions are also worth discussing. The intensity of both pain and happiness are high during episodes of sports and exercise, especially for men. This pattern, which is not surprising, may result from elevated endorphins during exercise. The low rating of "interested" during education-related activities might be related to the high dropout rate of college-age students in the United States. Telephone calls seem to evoke a high level of diverse emotions, with above-average ratings of happy, stressed, sad, and interest. Medical care is rated as an especially painful
46 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.8
U-Index and average of selected emotions by activity
ATUS activity category
U-index
No. of
(%)
Happy Stressed Sad Interested Pain episodes
Religious
6.4
4.97
0.90 0.66
5.09
0.61
151
Sports and exercise
7.4
5.08
0.84 0.25
4.92
1.20
321
Eating and drinking
9.7
4.57
1.11 0.52
4.03
0.80 1,206
Relaxing and Leisure
13.4
4.34
1.08 0.70
4.55
0.91 1,173
Socializing
13.5
4.74
1.21 0.66
4.65
0.88
528
Lawn and garden
14.2
4.23
0.98 0.47
3.92
1.37
318
Child care
15.6
4.63
1.76 0.39
4.41
0.56
376
Shopping
16.9
4.11
1.42 0.45
4.04
0.85
342
Volunteer
17.7
4.22
1.40 0.61
4.86
0.57
53
Watching TV
18.1
3.91
1.17 0.82
3.97
0.94 1,946
Food prep and clean-up
19.0
4.02
1.58 0.62
3.62
1.07
595
Travel
20.7
4.05
1.69 0.59
3.46
0.81 1,150
Telephone calls
23.5
4.47
2.02 1.14
4.99
0.86
128
Personal care
23.6
4.02
1.83 0.91
3.32
1.30
172
Housework
24.0
3.55
1.46 0.61
3.16
1.02
538
Working
26.9
3.80
2.37 0.69
3.99
0.71 1,671
Household management 27.9
3.50
1.85 0.82
3.94
0.76
235
Medical care
29.0
3.64
2.50 0.75
4.06
1.66
77
Education
32.3
3.62
2.66 0.87
3.87
0.82
143
Adult care
33.8
3.54
1.89 1.46
3.63
1.34
67
All
18.6
4.13
1.53 0.66
4.03
0.88 11,781
Source: Authors' calculations based on PATS. Notes: U-index indicates the proportion of fifteen-minute intervals in which stressed, sad, or pain exceeded happy.
activity, particularly by women. The emotional experience of watching television appears quite close to the overall average emotional experience during the day, except for stress, which is below average. A salutary feature of the PATS is that the same individual reports on multiple episodes of the day. As a result, individual fixed-effects (means) can be removed when studying differences in activities. Table 1.9 reports the U-index and affective ratings during the various activities after removing individual fixed effects. In essence, this analysis compares the emotional ratings of the same individual as he or she moves from one activity to another. In general, the activities are ranked similarly with or without fixed effects removed. The correlation between the U-index across activities in Table 1.8 and 1.9 is 0.93. The biggest movement occurs for medical care and personal care, both of which become less unpleasant when person-effects are removed, indicating that the people who tend to engage in these activities have a higher-than-average U-index during other episodes of the day. Because people tend to seek medical care when they are in pain or ill, this finding is quite plausible.
National Time Accounting: The Currency of Life 47
Table 1.9
U-index and average of selected emotions by activity after removing individual fixed effects
ATUS activity category
U-index
No. of
(%)
Happy Stressed Sad Interested Pain episodes
Religious
8.3
4.81
0.94 0.83
5.14
0.88
151
Eating and drinking
10.7
4.49
1.14 0.55
3.99
0.78 1,206
Sports and exercise
11.9
4.89
1.22 0.48
4.87
1.48
321
Socializing
13.0
4.68
1.21 0.59
4.65
0.84
528
Child care
13.6
4.59
1.44 0.49
4.49
0.65
376
Relaxing and leisure
15.1
4.35
1.24 0.68
4.49
0.88 1,173
Watching TV
15.7
4.00
1.16 0.71
4.01
0.77 1,946
Lawn and garden
16.7
4.21
1.21 0.55
3.92
1.25
318
Personal care
17.4
4.07
1.47 0.60
3.20
0.96
172
Food prep and clean-up
17.6
4.02
1.42 0.51
3.39
0.92
595
Shopping
18.0
4.15
1.68 0.63
4.01
0.92
342
Travel
19.8
4.06
1.62 0.63
3.64
0.89 1,150
Telephone calls
20.4
4.50
1.73 0.94
5.14
0.84
128
Volunteering
20.7
4.28
1.72 0.81
4.71
0.96
53
Medical care
22.6
3.76
2.20 0.83
4.52
1.22
77
Housework
25.6
3.56
1.57 0.68
3.11
1.08
538
Household management 27.4
3.70
1.68 0.78
4.00
0.76
235
Education
28.7
3.55
2.39 0.90
4.09
0.80
143
Working
28.8
3.83
2.34 0.78
4.09
0.89 1,671
Adult care
32.0
3.50
1.79 1.15
3.37
1.23
67
All
18.6
4.13
1.53 0.66
4.03
0.89 11,781
Source: Authors' calculations based on PATS. Notes: U-index indicates the proportion of fifteen-minute intervals in which stressed, sad, or pain exceeded happy.
Another feature of the PATS is that affect can be modeled before, during, and after participating in a specific activity. Figure 1.7 illustrates this point by showing the average rating of the emotion "happy" in relation to the occurrence of an episode involving sports or exercise. Specifically, we regressed the happiness rating on the number of minutes before or after an episode involving exercise with an interaction to allow for a different slope before and after exercise, for the subset of people who exercised on the interview day. The model was estimated both with and without person fixed effects. Time zero corresponds to the period of exercise. Especially in the model that removes person fixed effects, an inverse-V pattern is evident: Happiness rises as a period of exercise approaches and then decays afterwards. With more observations, a less constraining model could be estimated. Krueger and Mueller (2008) use the PATS data to compare the wellbeing of employed and unemployed individuals. Many previous studies have found that the unemployed are much less satisfied with their lives (e.g., Clark and Oswald 1994). The PATS data likewise show significantly lower average
48 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Fig. 1.7 Happiness rating before and after exercise, results of a linear spline life satisfaction and a significantly higher U-index for the unemployed than employed. The PATS data enable one to further ask: during which activities are the unemployed particularly unhappy or sad? The results indicate that the unemployed are particularly sad during time periods involving job searching and television viewing. 1.5.3 Interaction Partners The presence of others during an episode affects the pleasantness of the experience. Table 1.10 presents the U-index for men and women, disaggregated by who else was present during the episode. The tabulations do not control for other features of the episode, but the pattern is generally similar when we control for the activity engaged in during the episode as well as person fixed effects. For simplicity, we present the unadjusted results here. When people are alone, the U-index is higher than when they interact with others. The identity of the "others" matters, however. For men, the U-index is lower when friends and relatives are present. Spending time with coworkers is associated with a higher U-index for both men and women, primarily because work has a high incidence of negative emotions, particularly stress. Spending time with the boss makes the experience of work notably more unpleasant. The pattern for men and women is similar, except for the striking elevation in the U-index for women when it comes to spending time
National Time Accounting: The Currency of Life 49
Table 1.10 Alone Spouse Children Parents Friends Coworkers Boss/supervisor
U-Index by whom with based on PATS data
Men (%)
Women (%)
p-value for difference between men and women
18.3
21.9
15.8
15.3
10.2
17.7
7.2
27.1
11.8
12.8
25.9
27.5
46.9
30.5
0.033 0.808 0.034 0.025 0.792 0.615 0.522
with one's parents or children.26 These differences are partly explained by the different mix of activities that men and women engage in when they are with their parents and children. For example, men spend relatively more of their time with children watching television and traveling than do women, while women spend relatively more of their time with children engaged in child care and doing chores. Even holding activities constant, however, there are sizable differences in the U-index between men and women when they are in the company of their parents or children. 1.5.4 Day of Week Table 1.11 reports the U-index by day of the week (i.e., the diary day). A test of a constant U-index across days is rejected at the 0.01 level. Not surprisingly, weekend days are associated with less unpleasant feelings than weekdays, although the U-index is slightly lower on Fridays than on Saturdays. (For many people, apparently the weekend starts on Friday.) The U-index is lowest on Sundays and slightly higher on Mondays than on Tuesdays through Thursdays. Almost half of the weekend-weekday difference in the U-index can be accounted for by the different mix of activities that take place on the weekend. The empirical support for the song "rainy days and Mondays always get me down" thus far is limited, as a statistical test does not find the U-index on Monday to be significantly higher than on other weekdays (t 1.41), and the evidence on rain on the diary day cited in table 1.5 was mixed as well.27 1.5.5 Goods and Time Use In the standard economic model, people consume goods to increase their utility. Time-use data are notably lacking in information on goods 26. The ranking in Table 1.9 for women is exactly the same as was found for interaction partners in the Texas DRM, except parents were not separately identified in the DRM. 27. Stone et al. (1985) provide related evidence.
50 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.11
U-Index by day of week based on PATS data
(%)
Monday
21.7
Tuesday
19.0
Wednesday
20.9
Thursday
20.1
Friday
16.8
Saturday
17.7
Sunday
13.7
consumption. Instead, it can be hoped that the activity description reflects the goods consumed during an episode or that no goods are involved. In many situations, however, this is likely to be inadequate. For example, food must be involved during episodes of eating, but we lack information on the quantity or quality of food. Dinners at McDonalds' or the French Laundry are obviously not equivalent experiences, yet these events are lumped together in the time-use data. When computed at the episode level, the U-index potentially reflects features of the episode, such as consumption of goods, that are not captured elsewhere in the data. Unobserved features of activities, including goods consumption, surely account for some of the variability in emotional responses across respondents engaged in a given activity. The largest expenditure item for most people is their housing. Wong (2007) merged data on housing values and other housing characteristics to the Columbus DRM to explore the effect of housing consumption on subjective well-being. She finds that respondents who live in larger or more expensive homes do not report higher net affect while they are at home (either absolutely or in comparison to time spent away from home). This conclusion holds for both women with and without children living at home. She also finds that reported joy from one's house and home is unrelated to the market value of the home but is positively related to the market value of the homes in the neighborhood. To illustrate the effect of the consumption of goods on the affective experience of time use, in the PATS we collected information on the size of the television set being viewed during episodes of watching television. Because television absorbs such a large proportion of people's time, this seemed a particularly worthwhile activity to focus on. Specifically, we asked respondents whether the television screen they were watching was greater than or smaller than twenty-five inches. (If we were to redo the survey today, we might ask about flat screen versus not-flat screen.) We regressed each of the reported emotions during television watching on an indicator for the size of the television set, education, household income, and the mean affect rating during other episodes of the day. The results indicated some emo-
National Time Accounting: The Currency of Life 51 tional benefit from watching a larger television: stress was lower (t ­2.7) and net affect was higher (t 2.0) if a larger television was being watched. Although we would not make too much out of this result, it does suggest the utility of collecting information on the nature of the goods involved during participation in certain activities. Clearly more could be done in connecting goods to the quality of experiences. For example, the nature of kitchen equipment could be related to affect during episodes involving cooking, and the make and model of cars could be related to affect during episodes of travel. Note, however, that goods only affect people's hedonic experience when they attend to them. For example, Schwarz, Kahneman, and Xu (2009) explored how the quality of the car driven (as indexed by the car's Bluebook value) affects the driver's emotional experience. They found that drivers feel better driving luxury cars than economy cars--but only during episodes that are car-focused; that is, in the 2 percent of episodes that the drivers categorized as "driving for fun." In the other 98 percent of driving episodes, like commuting to work or shopping, the type of car driven was unrelated to drivers' emotional experiences. In short, the car only made a difference when the car was on the driver's mind. However, drivers are not aware of this contingency and drivers of luxury cars reported that they "generally" feel much better while driving than drivers of economy cars. Such discrepancies between global and episodic reports of enjoyment highlight that global reports of one's "usual" experience are based on general beliefs about the type of activity, which are often at odds with actual experience as captured by episodic assessments. 1.5.6 Decomposing Group Differences: The Case of Age and Income Age Past research finds that older individuals report fewer negative emotional experiences and greater emotional control than younger individuals (e.g., Gross et al. 1997). Consistent with this result, we find that the U-index is lower for those age sixty-five and older than for the younger population. The younger group works more and spends more time taking care of children, activities associated with stress (see fig. 1.8). How much of the difference in the U-index between young and old is accounted for by differences in their activities? Here we provide an example of how the difference in well-being between groups can be attributed to differences in time allocated across activities and differences in affect derived from a given set of situations and a residual. To simplify the analysis, we focus on the gap in the U-index between people age twenty-five to sixty-four and those sixty-five and over. We also confine our attention to weekdays, when differences in activities are more pronounced. Table 1.12 summarizes our results. The U-index is 20.4 percent
52 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Fig. 1.8 Time spent in various activities by age, 2005 ATUS
Table 1.12
Decomposition of U-index for 25 to 64-year-olds and those 65 and over
Group
Actual (%)
Predicted (%)
Unexplained by activities (%)
25­64-year-olds
20.4
20.0
0.4
65
16.1
17.5
­1.4
Difference
4.3
2.5
1.8
Notes: Table gives actual episode-level U-index and the predicted U-index using the overall sample's average U-index at the activity-level. Seventy-two harmonized activities are used.
for the younger group and 16.1 percent for the older group, a gap of 4.3 points ( p 0.007). If we compute the U-index using each group's actual time allocation and the average activity ratings for the combined sample (so the entire difference is due to differences in time allocated across activities), the gap is predicted to be 2.5 points.28 Thus, 58 percent ( 100 2.5/4.3) of the difference in the U-index between young and working-age is solely a result of differences in their activities. The remaining 1.8 point gap is a result of differences in emotional responses to the same set of activities or an interaction between differences in ratings and differences in time allocation. 28. This is mostly a result of the difference in working hours. During weekdays the younger group spent 24 percent of its awake time at work compared with just 2.6 percent for the older group. The U-index is 9 points higher during work-related episodes. So 9 percent (.24­.026) 1.9 points of the 2.5 points is due to the difference in time spent at work.
National Time Accounting: The Currency of Life 53
Table 1.13
Decomposition of U-index by income group
Group
Actual (%)
Predicted (%)
Unexplained by activities (%)
40,000
23.2
20.4
2.8
$75,000
19.0
21.4
­2.4
Difference
4.2
­1.0
5.2
Notes: Table gives actual episode-level U-index and the predicted U-index using the overall sample's average U-index at the activity-level. Seventy-two harmonized activities are used.
A further indication that choice of activity plays a role here comes from comparing the weekend and weekdays. On the weekend, the U-index falls to 16.8 percent for the younger group, not very different from the U-index for the older group during the week.29 Income Unlike the gap in U-index between older and younger groups, differences in time use across activities do not help explain the difference in U-index between income groups. To illustrate, we divided the sample of people age twenty-five to sixty-four into two groups, those in families with annual income less than $40,000 and those in families with annual income of $75,000 or more. Table 1.13 summarizes our results. The U-index is 23.2 percent for the lower income group and 19.0 percent for the higher income group, a gap of 4.2 points. If we recompute the U-index using each group's actual time allocation and the average activity ratings for the combined sample (so the entire difference is due to differences in time allocated across activities), the gap is predicted to be ­1 point. That is, the lower income group spends slightly more time in activities that are rated lower on the U-index. So the higher income group has a comparatively lower U-index because it rates the same activities as more enjoyable than does the lower income group. Episodes of TV watching, for example, have a lower U-index for the higher income group. One reason why differences in activities might explain a large share of the age gap in the U-index but not of the income-gap involves reverse causality. High-income earners may earn high incomes, in part, because they have cheerful personalities that enable them to prosper in the job market. Those who tend to be depressed and unhappy, on the other hand, are likely to suffer an income loss as a result. Causality runs, at least in part, from personality trait to income. Differences in personalities between income groups are likely to permeate their feelings throughout the day, regardless of the
29. The U-index also falls for the older group, but by a smaller amount, to 13.4 percent. Perhaps the elderly are more cheerful on the weekend because they interact with more cheerful younger people on those days.
54 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone activities individuals engage in. In contrast to income groups, personality differences between age groups are likely to be less important because age is exogenously determined. 1.6 Identifying Affectively Similar Activities Summarizing time-use data at the activity level can be unwieldy.30 The ATUS, for example, has hundreds of detailed activity codes. To make the analysis tractable, it is necessary to group activities into common categories. But classifying activities requires judgments of what activities are similar. Should gardening and lawn care be classified with leisure activities or with home production activities, for example? Researchers may have a different view of the enjoyment derived from such activities than the general public would. (See Aguiar and Hurst [2007] and Ramey and Francis [2006] for alternative results in which researchers classified time use into broad categories, such as leisure, home production, and market work. For results of an alternative approach that classifies leisure based on individuals' interactions with others, see Nadal and Sanz [2007]). Rather than externally assign activities to groups, we propose an alternative approach: use the average of the emotional ratings that respondents reported during each activity to assign activities with similar emotional experiences to the same group. Specifically, we use K-means cluster analysis to identify K groups of activities associated with similar emotional experiences. Cluster analysis is a family of techniques for assigning observations to groups (clusters) in a way that minimizes the discrepancies within groups and maximizes discrepancies between groups. For a single outcome measure (e.g., happy), the K-means cluster technique minimizes the withincluster variance, which also has the feature of maximizing the betweencluster variance in means. The interpretation is more complicated with more than one outcome measure, but the intuition is the same. The algorithm for the Stata cluster procedure used here minimizes the sum of squared Euclidean distances of the emotions associated with the activities from their cluster means (which is equivalent to maximizing between group differences as well due to a multivariate extension of the Pythagorean identity from analysis of variance [ANOVA]). We illustrate this approach using ratings of pain, happy, tired, stressed, sad, and interested to cluster activities. Activities form the unit of observation. For each activity, we computed the weighted average of each of those six emotional responses. Activities in the PATS were originally coded with the same system that the Census Bureau uses for ATUS. Because we will use the groups to make historical comparisons in section 1.7, we converted the ATUS activity codes to seventy-two "harmonized" codes used in the 30. This section and the next one borrow heavily from Krueger (2007).
National Time Accounting: The Currency of Life 55 American Heritage Time Use Studies (AHTUS).31 These harmonized codes are activity codes that can be compared over time in a consistent way. We set K to equal 6, mainly because 6 is a tractable number of categories and because it is not very different from the number of categories that researchers have used in the past. It would be possible to explore the sensitivity of the results to other values of K, or to select K on the basis of a goodness of fit test. Two additional features of the analysis are worth noting. First, the activities were weighted by their relative frequencies.32 Thus, the resulting clusters can be thought of as minimizing the weighted sum of within-group variances. Second, because cluster analysis is an iterative procedure that can be sensitive to the starting point, we executed the cluster command thirty-five times using random starting points and selected the estimates with the highest Calinski and Harabasz pseudo-F statistic, defined as: trace(B)/(g 1) F trace(W) /(n g) , where B is the between-cluster sum of squares and cross-products matrix, W is the within-cluster sum of squares and cross-products matrix, g is the number of groups, and n is the sample size. Table 1.14 reports the optimal cluster assignments for the most common activities and the average ratings for each of the six emotions. In addition, the table reports net affect, the positive emotion (happy) less the average of the negative ones (sad, pain, stressed). Many of the cluster assignments make intuitive sense. Paid work performed at home and away from home, for example, are both in cluster 6, as is helping someone with homework. Home production activities, including cleaning and putting away dishes, are mostly assigned together in cluster 5. There are some unexpected results, however. For example, time on a second job is classified in cluster 2 while other paid work is in cluster 6. Unfortunately, we did not collect occupation or industry for secondary jobs. Compared with surveyed episodes during the main job, people on a second job were much less likely to work with coworkers and were more likely to work alone or with their spouse. In addition to tracking and organizing time use, another application of the classification of activities that result from this exercise would be for nonmarket NIPAs. In particular, a question often arises in valuing nonmarket 31. The concordance was from the Center for Time Use Research (www.timeuse.org/ athus/documentation). The concordance contains ninety-two activities, fourteen of which could not be coded in the ATUS. We combined child care regardless of the child's age. We omitted sleeping and napping and a small number of infrequent activities that were not covered by PATS, resulting in seventy-two harmonized activities. 32. Because Stata does not have a weight option with cluster, we created a new data set in which each activity could be represented multiple times, in proportion to its relative frequency.
Table 1.14
Clusters assigned based on six emotions, 2006 PATS
Activity
Net
No. of
Cluster
effect
Happy
Tired
Stress
Sad
Interested
Pain
episodes
Personal medical care Financial/government services Homework Purchase medical services Writing by hand Purchase routine goods Other child care Use computer Second job, other paid work Other meals and snacks Walking General voluntary acts Conversation, phone, texting Read books Receive or visit friends Read to/with, talk with children Travel related to consumption Other in-home social, games Pet care, walk dogs Worship and religious acts Sports and exercise Cafй, bar
1
0.21
2.34
3.69
2.21
1.06
2.70
3.10
24
1
0.32
2.87
3.19
3.40
1.86
3.34
1.92
20
1
0.80
2.71
3.08
3.32
0.94
3.08
1.47
43
1
2.08
3.67
2.77
2.51
0.74
4.08
1.63
80
2
2.79
3.46
1.97
0.96
0.52
3.69
0.53
34
2
3.08
4.03
2.29
1.46
0.52
3.96
0.88
218
2
3.08
3.93
2.43
1.32
0.48
3.79
0.73
30
2
3.24
3.99
2.17
1.16
0.55
4.52
0.55
240
2
3.40
4.39
2.49
1.42
0.66
4.48
0.90
67
2
3.61
4.47
2.42
1.15
0.58
3.91
0.83
971
2
3.95
4.66
1.56
0.64
0.27
4.21
1.22
56
3
3.36
4.22
2.41
1.40
0.61
4.86
0.57
53
3
3.42
4.55
2.44
1.50
0.93
4.61
0.98
377
3
3.49
4.36
2.35
0.94
0.83
4.81
0.87
474
3
3.79
4.71
2.71
1.25
0.59
4.77
0.90
187
3
3.92
4.73
2.61
1.45
0.39
4.72
0.58
35
3
4.04
5.02
2.87
1.86
0.51
4.23
0.55
18
3
4.08
4.77
2.23
1.04
0.25
4.92
0.78
121
3
4.14
4.91
2.89
1.06
0.49
4.51
0.75
104
3
4.24
4.97
1.70
0.90
0.66
5.09
0.61
151
3
4.26
5.09
2.87
0.89
0.25
4.97
1.34
208
3
4.39
5.00
2.24
0.88
0.29
4.59
0.66
255
General out-of-home leisure
3
4.39
4.91
1.91
0.46
0.38
4.49
0.69
29
Purchase personal services
3
4.43
5.06
2.08
0.69
0.16
4.33
1.05
22
Parties or receptions
3
4.72
5.24
2.04
0.88
0.29
5.00
0.38
90
Hunting, fishing, boating, hiking
3
4.73
5.32
1.91
0.74
0.36
5.26
0.68
30
Attend sporting event
3
4.74
5.24
1.73
0.78
0.04
4.97
0.69
21
Play with children
3
4.81
5.41
2.49
0.74
0.21
4.69
0.86
40
Listen to music (cd, etc.)
3
4.81
5.33
1.56
0.38
0.35
5.06
0.84
22
Watch television, video
4
2.94
3.91
2.94
1.17
0.82
3.97
0.94
1,946
Food preparation, cooking
4
3.14
4.25
2.65
1.63
0.60
3.91
1.11
452
Relax, think, do nothing
4
3.25
4.40
2.77
1.31
0.80
3.96
1.34
313
Gardening
4
3.34
4.26
2.79
0.92
0.43
3.88
1.41
306
Set table, wash/put away dishes
5
2.28
3.32
2.81
1.45
0.68
2.76
0.93
145
Laundry, ironing, clothing repair
5
2.46
3.33
2.28
1.11
0.61
2.73
0.94
187
Adult care
5
2.56
3.90
2.56
1.72
1.19
3.82
1.10
87
Cleaning
5
2.63
3.72
2.85
1.61
0.62
3.54
1.05
327
Other domestic work
5
2.63
3.76
2.59
1.85
0.66
3.87
0.90
368
Travel related to leisure/other
5
3.00
4.02
2.73
1.66
0.57
3.43
0.79
1,120
Wash, dress, personal care
5
3.11
4.31
3.16
1.78
0.77
3.39
1.02
140
Home repairs, maintain vehicle
6
2.22
3.50
2.76
1.97
0.85
3.95
1.03
89
Paid work at home
6
2.35
3.47
2.66
2.01
0.63
4.00
0.71
207
Regular schooling, education
6
2.42
3.77
3.73
2.69
0.89
4.01
0.48
70
Main paid work (not at home)
6
2.55
3.83
2.72
2.44
0.69
3.98
0.71
1,425
General care of older children
6
3.55
4.54
3.41
1.98
0.45
4.36
0.54
235
58 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone activities whether an activity should be valued at the wage rate, at the market wage for hiring someone to perform a task, or at some other price. Another issue concerns whether particular activities such as schooling are primarily consumption activities or investment activities. One answer to this question is that activities that are as stressful and uninteresting as someone's main job should be valued at the same wage as the main job. Likewise, activities that are as enjoyable as socializing should be treated as leisure. The cluster analysis provides a means for identifying activities that are associated with similar emotional experiences. For example, time spent in school does not appear to be a consumption activity in our data, and time spent taking care of teenagers appears as taxing as one's main job. Table 1.15 reports the mean of the emotions and net affect for each cluster of activities. The lowest rated cluster in terms of net affect is cluster 1, which includes receiving medical care, purchasing medical services, seeking government services, and doing homework. Cluster 2 involves tasks like writing and using a computer. The most enjoyable and interesting activities are in cluster 3, including religious activities, exercise, attending parties, listening to music, playing with children, and recreation. Cluster 4 is a mixture of activities, such as watching television, relaxing, cooking, and gardening, that are close to average in terms of affect ratings. Cluster 5, which includes domestic activities such as doing laundry, ironing, caring for adults, and cleaning, is slightly above cluster 6 (work) in terms of net affect but well below it in terms of interest. If we were to assign value-laden terms to describe the clusters, we could think of cluster 1 as unpleasant personal maintenance, cluster 2 as moderately enjoyable tasks, cluster 3 as engaging leisure and spiritual activities, cluster 4 as neutral downtime and cooking, cluster 5 as mundane chores, and cluster 6 as work-like activities. One caveat to bear in mind is that average affect ratings are conditional on engaging in the activity for a given length of time. People probably sort into the activities that they engage in based, in part, on how much utility
Table 1.15
Average of emotions by cluster
Cluster Happy Tired Stressed Sad Interested Pain Net effect
1
3.09
2.97
2.92
1.18
3.57
1.80
1.12
2
4.29
2.31
1.18
0.55
4.06
0.78
3.45
3
4.79
2.37
1.05
0.56
4.79
0.84
3.97
4
4.05
2.87
1.23
0.76
3.95
1.06
3.04
5
3.86
2.72
1.64
0.63
3.44
0.89
2.80
6
3.88
2.83
2.35
0.69
4.04
0.69
2.63
Notes: Averages are weighted by episode frequency and sample weights. All emotions are reported on a 0 to 6 scale. Sample is PATS data. Based on July 5, 2007, cluster6_freqwgt_ctus_ best.log.
National Time Accounting: The Currency of Life 59 they derive from them. If the cluster analysis is redone using residuals of the six emotions after removing person effects, however, 83 percent of activities (weighted by frequency) remain in the same cluster as in the original assignment that did not remove person effects. Thus, the cluster analysis seems to provide a reasonably robust and plausible set of groups of activities that can be used to compare time use over time or between countries. 1.7 Comparing Time Use over Time in Groups of Activities and Generally We propose three techniques for tracking time use over time: (a) following groups of activities defined in section 1.6, (b) computing an overall U-index based on the U-index associated with various activities at a point in time; and (c) computing the U-index at the episode level. To illustrate the first two techniques, we used data from a project originally of the Yale University Program on Nonmarket Accounts, known as the American Heritage Time Use Studies (AHTUS). The AHTUS consists of five time-use surveys conducted from 1965 and 1966 through 2003. The disparate activity codes were harmonized to a common set of seventy-two main activities (plus missing/unclassified). In addition, we merged the harmonized activity codes to the 2005 ATUS and include it as well. The underlying sources of the harmonized data are described in the following box. Unfortunately, it is not possible to compute the episode-level U-index over time as PATS-like data are not available in earlier years, so we just illustrate the technique. We hope that data will be available in the future for episode-level analyses. Historical Time-Use Surveys · 1965­1966: Original source is Multinational Comparative TimeBudget Research Project conducted by the University of Michigan's Survey Research Center. N 1,968. · 1975­1976: Original source is American's Use of Time: Time Use in Economic and Social Accounts, conducted by the University of Michigan's Survey Research Center. N 5,869. · 1985: Original source is American's Use of Time, conducted by the University of Maryland's Survey Research Center. N 2,308. · 1992­1994: Original source is National Human Activity Pattern Survey, conducted by the University of Maryland's Survey Research Center. N 5,964. · 2003: Original source is ATUS, conducted by Census Bureau for Bureau of Labor Statistics. N 15,999. · 2005: Original source is ATUS, conducted by Census Bureau for Bureau of Labor Statistics. N 10,112.
60 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Sample weights were used for all estimates using the AHTUS data sets. Because we lack affect ratings during sleep, we focus on the waking day.33 One issue that we can only partially address is that the data sets use different methods and sampling frames. For example, the 1965 to 1966 survey sampled people from households in which someone was employed in a nonagricultural industry, and only covered certain months of the year. The samples were restricted to those from age nineteen to sixty-four to have a consistent age range. The average age was fairly similar in the data sets, ranging from 38.4 in 1985 to 40.6 in 2003. 1.7.1 Tracking Groups of Activities Table 1.16, panels A and B, present the average proportion of women's and men's awake time spent in the harmonized activities, respectively. A motivation of the cluster analysis was to classify these activities into affectively similar categories so that changes in time use could be tracked in a more manageable set of categories. Specifically, for each person we first computed the average percentage of the awake day spent in each of the six clusters previously described. We next averaged over every individual in the sample.34 Table 1.17, panel A, summarizes the results for men and women combined. The picture that emerges is one of stability for clusters 1 (unpleasant personal maintenance), 2 (moderately enjoyable tasks), and 6 (work-like activities). Time spent on cluster 4 (neutral downtime) is up while cluster 3 (engaging leisure) and cluster 5 (mundane chores) are down. Overall, these figures suggest that affectively neutral downtime activities like watching television have gained at the expense of mundane chores and engaging leisure activities over the last forty years. Panels B and C of table 1.17 report separate results for men and women, respectively. For men, the share of the day devoted to cluster 6 (work-like activities) has declined by 6 percentage points since 1965 and 1966, while the share devoted to cluster 4 (neutral downtime) has increased by 8.5 points. Women, not surprisingly, have increased time in cluster 6 activities by 5 percentage points because of higher labor force participation, while time spent on mundane chores fell even more, by almost 7 points. The amount of time women spend in cluster 3 (engaging leisure) fell by roughly the same amount (3 points) as their time devoted to cluster 4 (neutral downtime) increased. These shifts, on balance, do not suggest significant improvements in affective experience for women over this entire forty-year time span. 33. Sleep rose from 7.95 hours in 1965 and 1966 to 8.5 hours in 2005, or by 2.3 percentage points on a twenty-four hour day. 34. Because a small number of activities (accounting for less than 3 percent of awake time each year) were not assigned to clusters in the PATS, they are omitted here. The percentages were renormalized to sum to 100 percent accordingly.
Table 1.16
Percentage of days spent in each activity, 1965­1966 to 2005
Main Activity
1965­ 1966 (%)
1975­ 1976 (%)
1985 (%)
1992­ 1994 (%)
1 General or other personal care 2 Wash, dress, personal care 3 Personal medical care 4 Meals at work 5 Other meals and snacks 6 Main paid work (not at home) 7 Paid work at home 8 Second job, other paid work 9 Work breaks 10 Other time at workplace 11 Time looking for work 12 Regular schooling, education 13 Homework 14 Short course or training 15 Other education or training 16 Food preparation, cooking 17 Set table, wash/put away dishes 18 Cleaning 19 Laundry, ironing, clothing repair 20 Home repairs, maintain vehicle 21 Other domestic work 22 Purchase routine goods 23 Purchase consumer durables 24 Purchase personal services 25 Purchase medical services 26 Purchase repair, laundry services 27 Financial/government services 28 Purchase other services 29 General care of older children 30 Medical care of children 31 Play with children 32 Supervise/help with homework 33 Read to/with, talk with children 34 Other child care 35 Adult care 36 General voluntary acts 37 Political and civic activity 38 Worship and religious acts 39 General out-of-home leisure 40 Attend sporting event 41 Theater, concert, opera 42 Museums, exhibitions 43 Cafй, bar 44 Parties or receptions 45 Sports and exercise
A. Women 1.52 0.20 5.80 4.90 0.06 0.11 0.74 0.69 7.09 7.83 14.32 14.07 0.62 0.56 0.14 0.17 0.51 0.34 0.23 0.19 0.00 0.08 0.19 0.30 0.30 0.42 0.21 0.20 0.72 0.03 7.46 7.08 3.71 2.26 5.94 5.76 4.43 2.45 0.30 0.60 1.58 0.59 1.90 2.94 0.14 0.12 0.27 0.26 0.13 0.25 0.33 0.16 0.06 0.14 1.52 0.10 3.47 2.36 0.09 0.12 0.32 0.30 0.25 0.13 0.24 0.36 0.30 0.57 0.67 1.10 0.45 0.29 0.09 0.04 0.95 1.09 0.16 0.18 0.11 0.26 0.02 0.09 0.01 0.04 0.11 0.27 1.54 0.55 0.34 0.60
0.79 6.67 0.04 0.72 7.32 15.83 1.36 0.26 0.18 0.16 0.08 0.33 0.48 0.28 0.16 5.77 1.87 4.52 1.99 0.40 1.49 3.10 0.08 0.16 0.30 0.10 0.20 0.19 2.23 0.07 0.41 0.16 0.18 0.23 0.51 0.43 0.01 0.84 0.16 0.28 0.06 0.01 0.49 0.55 0.98
0.32 5.84 0.06 0.00 6.88 21.10 0.81 0.01 0.06 0.00 0.06 1.01 0.77 0.04 0.09 4.09 0.68 4.79 1.58 0.39 1.40 0.93 2.60 0.18 0.37 0.09 0.12 0.10 1.44 0.02 0.33 0.18 0.06 0.43 0.51 0.05 0.00 1.02 0.00 0.31 0.14 0.06 0.30 0.69 1.50
2003 (%)
2005 (%)
0.25 0.09 5.22 4.96 0.44 0.64 0.05 0.03 5.27 5.51 19.51 19.13 1.36 1.28 0.64 0.62 0.02 0.02 0.00 0.00 0.18 0.14 0.61 0.43 0.79 0.70 0.06 0.21 0.02 0.02 3.74 3.77 1.23 1.22 3.97 4.58 2.21 2.37 0.32 0.28 1.26 1.24 3.35 3.31 0.01 0.02 0.26 0.19 0.43 0.33 0.12 0.11 0.09 0.10 0.06 0.06 2.60 2.37 0.16 0.17 0.87 0.81 0.52 0.45 0.38 0.43 0.54 0.53 1.65 1.35 0.91 0.78 0.02 0.00 0.98 0.89 0.19 0.21 0.22 0.16 0.11 0.08 0.06 0.05 1.63 1.44 0.68 0.61 0.90 0.84 (continued )
Table 1.16
(continued)
Main Activity 46 Walking 47 Cycling 48 Physical activity/sports with child 49 Hunting, fishing, boating, hiking 50 Gardening 51 Pet care, walk dogs 52 Receive or visit friends 53 Other in-home social, games 54 Artistic activity 55 Crafts 56 Hobbies 57 Relax, think, do nothing 58 Read books 59 Listen to music (cd, etc.) 60 Listen to radio 61 Watch television, video 62 Writing by hand 63 Conversation, phone, texting 64 Use computer 65 Imputed travel 66 Travel related to personal care 67 Travel related to work 68 Travel related to education 69 Travel related to consumption 70 Travel related to child care 71 Travel related to volunteering/worship 72 Travel related to leisure 73 Missing/unclassified 1 General or other personal care 2 Wash, dress, personal care 3 Personal medical care 4 Meals at work 5 Other meals and snacks 6 Main paid work (not at home) 7 Paid work at home 8 Second job, other paid work 9 Work breaks 10 Other time at workplace 11 Time looking for work 12 Regular schooling, education 13 Homework 14 Short course or training 15 Other education or training 16 Food preparation, cooking 17 Set table, wash/put away dishes 18 Cleaning
1965­ 1966 (%)
1975­ 1976 (%)
0.10 0.00 0.05 0.08 0.27 0.13 4.97 0.46 0.07 1.24 0.04 0.59 3.02 0.08 0.28 8.47 0.74 1.60 0.00 0.00 0.71 1.35 0.11 2.13 0.55 0.39 1.89 1.34 B. Men 0.93 4.60 0.06 1.55 7.49 34.98 0.97 0.96 1.16 0.68 0.00 0.32 0.73 0.26 0.29 0.84 0.35 0.94
0.13 0.03 0.13 0.21 0.55 0.37 4.78 0.69 0.15 1.44 0.04 1.16 2.97 0.20 0.19 12.74 0.23 2.20 0.00 0.05 0.96 1.37 0.13 2.06 0.53 0.91 1.87 2.79 0.19 4.04 0.04 1.18 8.42 30.28 1.76 0.71 0.60 0.40 0.16 0.67 0.76 0.25 0.09 1.03 0.22 1.79
1985 (%) 0.25 0.02 0.15 0.25 0.36 0.57 2.94 0.71 0.11 0.76 0.02 0.74 2.68 0.08 0.23 13.02 0.39 3.37 0.08 0.00 0.86 1.97 0.22 2.33 0.53 0.67 2.04 2.18
1992­ 1994 (%) 0.00 0.00 0.10 0.00 0.26 0.44 4.01 0.56 0.09 0.55 0.03 1.81 2.44 0.04 0.11 14.87 0.72 1.42 0.26 0.00 1.76 2.26 0.23 2.22 0.36 0.37 2.00 1.66
0.74 4.93 0.02 0.90 7.63 25.57 2.62 0.54 0.27 0.35 0.12 0.64 0.93 0.20 0.12 1.44 0.38 2.13
0.34 4.10 0.04 0.00 7.13 29.27 1.23 0.06 0.08 0.00 0.10 1.23 0.93 0.03 0.07 1.52 0.14 2.54
2003 (%) 0.31 0.03 0.02 0.08 0.82 0.60 4.62 0.58 0.02 0.11 0.02 1.77 1.96 0.10 0.07 13.60 0.19 0.92 0.89 0.33 1.56 1.68 0.13 2.50 0.77 0.27 1.71 0.47
2005 (%) 0.26 0.02 0.04 0.10 0.80 0.65 1.81 0.80 0.02 0.17 0.03 1.69 2.15 0.07 0.11 14.68 0.15 3.45 1.00 0.03 0.97 1.66 0.11 1.26 0.72 0.26 1.56 2.92
0.25 3.67 0.31 0.05 5.55 28.44 1.54 1.00 0.03 0.00 0.30 0.64 0.68 0.03 0.04 1.42 0.33 1.88
0.17 3.51 0.60 0.06 5.93 27.41 1.89 0.96 0.03 0.00 0.15 0.50 0.90 0.09 0.00 1.42 0.30 1.89
Table 1.16
(continued)
Main Activity 19 Laundry, ironing, clothing repair 20 Home repairs, maintain vehicle 21 Other domestic work 22 Purchase routine goods 23 Purchase consumer durables 24 Purchase personal services 25 Purchase medical services 26 Purchase repair, laundry services 27 Financial/government services 28 Purchase other services 29 General care of older children 30 Medical care of children 31 Play with children 32 Supervise/help with homework 33 Read to/with, talk with children 34 Other child care 35 Adult care 36 General voluntary acts 37 Political and civic activity 38 Worship and religious acts 39 General out-of-home leisure 40 Attend sporting event 41 Theater, concert, opera 42 Museums, exhibitions 43 Cafй, bar 44 Parties or receptions 45 Sports and exercise 46 Walking 47 Cycling 48 Physical activity/sports with child 49 Hunting, fishing, boating, hiking 50 Gardening 51 Pet care, walk dogs 52 Receive or visit friends 53 Other in-home social, games 54 Artistic activity 55 Crafts 56 Hobbies 57 Relax, think, do nothing 58 Read books 59 Listen to music (cd, etc.) 60 Listen to radio 61 Watch television, video 62 Writing by hand 63 Conversation, phone, texting 64 Use computer
1965­ 1966 (%) 0.11 0.99 0.79 1.05 0.18 0.09 0.17 0.25 0.04 1.02 0.40 0.00 0.46 0.08 0.06 0.11 0.47 0.21 0.10 0.59 0.03 0.14 0.05 0.02 0.66 1.40 0.72 0.16 0.00 0.04 0.52 0.16 0.06 3.29 0.54 0.11 0.01 0.28 0.31 3.46 0.10 0.44 11.21 0.27 0.99 0.00
1975­ 1976 (%) 0.10 1.75 0.72 1.31 0.15 0.05 0.14 0.13 0.13 0.11 0.48 0.02 0.17 0.05 0.11 0.13 0.91 0.24 0.02 0.76 0.08 0.30 0.08 0.05 0.48 0.59 1.24 0.19 0.03 0.07 0.63 0.38 0.34 3.36 0.52 0.05 0.22 0.32 1.21 2.61 0.42 0.28 12.77 0.12 1.53 0.00
1985 (%) 0.26 1.80 1.35 1.69 0.10 0.06 0.19 0.15 0.16 0.23 0.38 0.01 0.23 0.04 0.08 0.06 0.54 0.26 0.00 0.54 0.19 0.28 0.09 0.03 0.83 0.61 1.75 0.26 0.03 0.10 0.99 0.61 0.52 2.50 0.51 0.09 0.03 0.30 0.77 2.42 0.13 0.33 14.55 0.23 2.05 0.17
1992­ 1994 (%) 0.30 1.64 1.13 0.44 1.24 0.04 0.21 0.18 0.10 0.10 0.25 0.00 0.20 0.05 0.07 0.15 0.40 0.10 0.03 0.65 0.00 0.40 0.06 0.03 0.78 0.61 2.21 0.00 0.00 0.04 0.00 0.33 0.40 3.60 0.51 0.03 0.04 0.04 1.74 2.44 0.08 0.24 16.41 0.60 0.73 0.58
2003 (%)
2005 (%)
0.42 0.45 1.49 1.47 0.88 0.84 2.17 1.95 0.03 0.01 0.06 0.06 0.24 0.28 0.13 0.11 0.08 0.07 0.05 0.04 0.83 0.84 0.05 0.01 0.60 0.54 0.23 0.17 0.12 0.12 0.25 0.25 1.22 1.13 0.72 0.67 0.00 0.05 0.74 0.57 0.22 0.17 0.26 0.29 0.09 0.16 0.06 0.01 1.67 1.65 0.62 0.52 1.39 1.36 0.23 0.22 0.05 0.07 0.04 0.07 0.53 0.50 1.39 1.64 0.45 0.47 3.86 1.63 1.00 1.06 0.02 0.00 0.18 0.13 0.04 0.06 1.75 1.93 1.55 1.44 0.26 0.32 0.12 0.13 16.08 17.25 0.12 0.11 0.44 2.69 1.24 1.25 (continued )
64 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.16
(continued)
Main Activity 65 Imputed travel 67 Travel related to work 68 Travel related to education 69 Travel related to consumption 70 Travel related to child care 71 Travel related to volunteering/worship 72 Travel related to other purposes 73 Missing/unclassified Note: Based on PATS data.
1965­ 1966 (%) 0.00 3.68 0.19 1.63 0.28 0.37 2.06 1.60
1975­ 1976 (%) 0.04 3.19 0.27 1.41 0.21 0.81 1.97 2.67
1985 (%) 0.01 3.45 0.17 1.86 0.23 0.62 2.58 2.00
1992­ 1994 (%) 0.00 3.35 0.22 1.59 0.11 0.35 2.35 2.23
2003 (%) 0.24 2.86 0.15 2.12 0.32 0.24 1.79 0.47
2005 (%) 0.03 2.69 0.09 0.95 0.26 0.18 1.71 2.47
Table 1.17 Cluster 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6
Average percent of day by cluster, 1965­1966 to 2005
1965­1966 (%)
1974­1975 (%)
1985 (%)
1992­1994 (%)
2003 (%)
Panel A: All
4.2
3.6
3.9
5.8
4.4
10.7
12.1
11.8
9.5
11.1
19.8
19.6
19.0
16.5
18.3
16.3
20.3
20.1
21.2
20.6
17.6
15.2
16.3
14.6
14.0
31.4
29.2
28.9
32.4
31.6
Panel B: Men
4.5
4.0
4.2
5.0
3.9
10.7
11.5
11.2
9.4
10.8
18.2
17.5
17.8
15.5
17.4
14.5
17.3
18.8
20.7
20.9
9.7
10.2
12.6
11.4
10.4
42.4
39.5
35.4
38.0
36.5
Panel C: Women
4.0
3.2
3.6
6.5
4.9
10.7
12.5
12.3
9.6
11.3
21.2
21.5
20.2
17.3
19.2
17.9
23.0
21.3
21.6
20.2
24.7
19.6
19.6
17.2
17.5
21.5
20.1
23.0
27.8
26.9
2005 (%) 3.8 11.5 17.1 22.3 14.1 31.2 3.6 11.1 16.1 23.0 10.2 36.0 3.9 11.9 18.1 21.7 17.9 26.5
1.7.2 Activity-Based U-Index In addition to classifying and tracking time use in categories, it is useful to summarize time allocation in a single welfare measure. The U-index can be used for this purpose. As before, the U-index measures the percent of moments spent in an unpleasant state during each activity, where an
National Time Accounting: The Currency of Life 65 unpleasant state is defined as one where a negative emotion (sad, stress, or pain) strictly dominates the positive emotions (happy in this case). Specifically, we first computed the U-index for each harmonized activity using the 2006 PATS data for a pooled sample of men and women. For example, the U-index during paid work was 27 percent, during exercise it was 8 percent, and during television viewing it was 18 percent. We next computed the weighted average U-index where the weights were the percent of aavwearkaegetiUm-eitnhdeeaxv,edreangoetpeedrsUotn, esapcehntyienareaisc:h activity. Formally, the weighted Ut iwit(iw jpitijtUj) , where wit is the sample weight for individual i, pijt is the proportion of time individual i spent in activity j in year t, and Uj is the U-index for activity j from the PATS. Panel A of table 1.18 reports the results. The activity-based U-Index shows very little trend over the last forty years for men and women combined or for women as a group. For men, however, there has been a shift away from activities associated with unpleasant feelings. To put the estimates in context, note that the difference between the activity-based U-index on weekends and weekdays is about 3 percentage points.35 Thus, the 1 point drop in the U-index from 1965 and 1966 to 2005 is about one-third of the difference in unpleasant feelings associated with activities during the week and those on the weekend. Although the U-index is highly correlated across activities for men and women, there are some notable differences in a small number of activities. Women, for example, find supervising/helping with homework and voluntary acts less unpleasant than do men. Thus, we computed the U-index separately for men and women. We then assigned the gender-specific U-index for each activity to each observation in the historical sample, and computed the activity-level U-index separately for men and women. Panel B of table 1.18 and figure 1.9 display the results, combining 2003 and 2005 for presentation. The results are generally consistent with those in panel A, though they are noisier. The gender-specific weighted U-index displays no trend for women and has trended downward for men over the last forty years, indicating an improvement in daily experience. Table 1.19 presents regressions to control for possible changes in the age and education composition of the samples, as well as the survey day and month. The unit of observation for the regressions is an individual. The dependent variable is the duration-weighted U-index for each person's activities on the survey day, or j pijt Uj, where Uj is the U-index for activity 35. With episode-level data, the weekend-weekday difference is about twice as large.
66 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.18
U-index based on time in various activities each year
1965­1966
1975­1976
1985
1992­1994
2003
2005
(%)
(%)
(%)
(%)
(%)
(%)
All Men Women Men Women
A. U-index from men and women combined
20.1
19.5
19.5
20.0
20.9
20.4
20.1
20.2
19.4
18.7
19.0
19.8
B. Gender-specific U-indices and time allocation
20.2
20.1
19.2
18.8
20.8
19.4
20.0
21.0
19.3
19.6
19.6
19.9
19.2
19.4
18.7
19.0
20.1
20.4
Note: A small number of missing and unclassified activities were assigned the mean U-index each year.
Fig. 1.9 Activity-level U-index over time, using gender-specific U-indexes j for men and women combined. The regression-adjusted estimates reveal a similar pattern: very little shift toward or away from unpleasant activities, on net, for women, but about a 1 percentage point shift away from activities associated with unpleasant feelings for men since the mid-1960s. Dispersion in Activity-Level U-Index The activity-level U-index masks some important trends across people and groups. The standard deviation of the activity-level U-index was calculated across people each year (see fig. 1.10). This measure of dispersion has grown by about 15 percent over the forty-year period. Thus, the spread in time people spend in activities according to their frequency of unpleasant moments is increasing over time.
Table 1.19
Regression models for activity-based U-index
All
Men
Women
Standard
Standard
Standard
Coefficient
error
Coefficient
error
Coefficient
error
Intercept Year 1975­1976 Year 1985 Year 1992­1994 Year 2003 Year 2005 Tuesday Wednesday Thursday Friday Saturday Sunday February March April May June July August September October November December Age Age-Squared Female HS Some college College College R2 Sample Size
20.905 ­0.518 ­0.544 ­0.031 ­0.682 ­0.409 ­0.137 0.007 ­0.194 ­0.513 ­2.231 ­3.018 0.022 0.203 0.056 ­0.118 ­0.146 ­0.406 ­0.405 ­0.018 0.088 0.142 0.102 0.036 ­0.001 ­0.921 ­0.048 0.438 0.152 0.009 0.104 40,388
0.224 0.074 0.070 0.071 0.070 0.070 0.071 0.071 0.071 0.071 0.071 0.072 0.089 0.092 0.095 0.093 0.089 0.111 0.107 0.096 0.095 0.087 0.089 0.011 0.000 0.038 0.059 0.052 0.056 0.075
21.108 ­0.338 ­0.731 ­0.677 ­1.255 ­0.950 ­0.122 0.035 ­0.049 ­0.553 ­2.599 ­3.431 ­0.128 ­0.072 ­0.179 ­0.272 ­0.302 ­0.351 ­0.473 ­0.221 0.028 ­0.031 0.082 0.054 ­0.001 -- ­0.025 0.511 0.103 ­0.006 0.115 17,921
0.356 0.118 0.111 0.113 0.110 0.109 0.113 0.113 0.112 0.112 0.113 0.113 0.140 0.146 0.149 0.146 0.142 0.177 0.171 0.152 0.150 0.140 0.140 0.017 0.000 0.093 0.084 0.087 0.112
19.862 ­0.689 ­0.363 0.551 ­0.130 0.110 ­0.149 ­0.023 ­0.325 ­0.474 ­1.893 ­2.645 0.158 0.451 0.243 0.004 ­0.018 ­0.470 ­0.363 0.177 0.109 0.313 0.092 0.018 0.000 -- ­0.113 0.329 0.142 ­0.054 0.084 22,467
0.279 0.094 0.088 0.089 0.090 0.089 0.090 0.090 0.090 0.090 0.090 0.090 0.113 0.115 0.121 0.117 0.112 0.139 0.134 0.121 0.120 0.109 0.113 0.013 0.000 0.074 0.066 0.072 0.099
Notes: Dependent variable is the duration-weighted average U-index. Regressions are estimated by weighted least squares. Person weights have been normalized to sum to one in each sample. Weighted mean (and standard deviation) of the dependent variable is 19.7 percent (4.0) for all, 20.1 percent (4.3) for men and 19.3 percent (3.8) for women. All explanatory variables are dummy variables except age and age-squared. Base year is 1965­1966. Dashed cells indicate there is no coefficient, since the gender variable is a constant for women and men.
68 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Fig. 1.10 Dispersion of activity-level U-index across people, 1965­1966 to 2003­2005 Additionally, the U-index has declined by more for men with a high school degree or less schooling than it has for men with a college degree or higher (see fig. 1.11). This result is consistent with Aguiar and Hurst's (2007) finding that leisure time increased more for the less educated than highly educated, partially offsetting the rise in income associated with additional schooling. 1.7.3 Episode-Level U-Index Table 1.7 provides what we refer to as episode-level estimates of the U-index for various groups. These are tabulations of the proportion of time spent in an unpleasant state where the episode is the unit of observation. The calculations do not require information on activities. If the nature of activities changes over time, the episode-level U-index will reflect this change. The episode-level U-index will also reflect the presence of others during the episode and other features of the episode. Moreover, if the U-index is calculated at the episode level, it allows for the fact that some people may respond emotionally to the same activity in different ways. Because activity and other measured features of episodes account for a small proportion of variability in affect--for example, controlling for seventy-one activity dummies only accounts for 6 percent of the variability in reported happiness across episodes--tracking changes over time in the episode-level U-index can be more informative than tracking how changes in activities are likely to affect well-being. Unfortunately, an episode-level U-index--either for a representative national sample or for selected groups--can only be calculated for 2006 because the PATS data set is cross-sectional. Nevertheless, the PATS data
National Time Accounting: The Currency of Life 69 Fig. 1.11 U-index for men, by education, 1965­1966 to 2003­2005 provide proof of the applicability of the idea and a baseline against which future measurements can be compared. If the affect questions are added to subsequent time-use surveys, such as ATUS, then the episode-level U-index can be computed at regular intervals in the future. 1.8 International Comparison In addition to comparing subjective well-being over time, social scientists and policymakers have long been interested in comparing SWB across countries.36 This interest partly stems from a desire to rank countries based on SWB. Additionally, cross-country data have been used to study the effect of various public policies, economic conditions, and institutions (e.g., Blanchflower 2007; Alesina, Glaeser, and Sacerdote 2002; Frey and Stutzer 2002). The most common measures of SWB in these studies are reports of overall life satisfaction or happiness, which reflect global evaluations of one's life relative to some standard. In this section, we compare SWB in two "representative" cities, one in France and the other in the United States, and ask whether the standard measure of life satisfaction and the DRM yield the same conclusion concerning relative well-being. Specifically, we designed a survey to compare overall life satisfaction, time use, and recalled affective experience during episodes of the day for random samples of women in Rennes, France and Columbus, Ohio. These cities were selected because they represent "middle America" and "middle France." We also present results using time allocation derived from national samples in the United States 36. This section is based on work that we did together with Claude Fischler. For a more detailed report see Krueger et al. (2009).
70 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone and France to extend our analysis beyond these two cities. This comparison illustrates national time accounting in a cross-national context. To preview the main results, based on the standard life satisfaction question, we find that Americans report higher levels of life satisfaction. Yet based on the DRM we find that the French spend their days in a more positive mood, on average. Moreover, the national time-use data indicate that the French spend relatively more of their time engaged in activities that tend to yield more pleasure than do Americans. Our results suggest that considerable caution is required in comparing standard life satisfaction data across populations with different cultures. In particular, the Americans seem to be more emphatic when reporting their well-being. The U-index apparently overcomes this inclination. 1.8.1 Study Design The sample consists of 810 women in Columbus, Ohio and 820 women in Rennes, France. They were invited to participate based on random-digit dialing in the spring of 2005. Respondents were paid approximately $75 for their participation in both countries. The age range spanned eighteen to sixty-eight, and all participants spoke the country's dominant language at home. The Columbus sample was older (median age of forty-four versus thirty-nine), more likely to be employed (75 percent versus 67 percent) and better educated (average of 15.2 years of schooling years versus 14.0) than the Rennes sample. In addition, the Rennes sample was more likely to be currently enrolled in school (16 percent versus 10 percent). The differences in demographic characteristics partly reflect different circumstances in the countries (e.g., the employment rate is 8 percentage points higher in the United States than in France, and average education is 0.9 years higher in the United States), and partly reflect idiosyncrasies of our two cities and sample. Because we compare SWB measured with different methods for the same samples, our results should reflect differences in the methods, not demographic differences between the samples. Essentially the same protocols as those used in the Texas DRM were followed. Groups of participants were invited for a weekday evening to a central location, where they completed a series of questionnaires contained in separate packets. The first packet included general satisfaction and demographic questions. The wording of the life satisfaction question closely followed the World Values Survey (although we use a different response set). The second packet asked respondents to construct a diary of the previous day as a series of episodes, noting the content and the beginning and ending time of each.37 The average number of episodes described was 13.2 in Columbus and 14.5 in Rennes. 37. About 300 participants in each country were recruited for Mondays to describe a weekend day. Half of them were instructed to describe the preceding Saturday and half the preceding Sunday. Data were not collected pertaining to Fridays.
National Time Accounting: The Currency of Life 71 In the third packet, respondents completed a form for each of the episodes they had previously listed. The form included a list of twenty-two activities and eight interaction partners, with an instruction to mark all that apply. Respondents who had checked multiple activities were requested to indicate the one that "seemed the most important to you at the time" (we call it focal). Unless specifically noted, all analyses refer to focal activities. The form also requested ratings of ten emotions that were experienced at the time on a scale from zero (not at all) to six (very strongly). We focus on the following emotions: "happy," "tense/stressed," "depressed/blue," and "irritated/angry." The questionnaire was translated back and forth between French and English to ensure common meanings, and some questions were modified and deleted as a result of this procedure. The data were reweighted by day of week to be representative of a random day. Weekdays received 5/7th of the weight and Saturday and Sunday received 1/7th of the weight in the weighted samples. Additional details of the procedures and all questionnaires are available online.38 1.8.2 Life Satisfaction Table 1.20 contains tabulations of reported life satisfaction in the two cities. As in most populations, reports of being very unsatisfied are rare. The American women, however, are twice as likely to say they are very satisfied with their lives as are the French women (26 percent versus 13 percent). Furthermore, assigning a number from one to four indicating life satisfaction, a common practice, also indicates that the Americans are more satisfied, on average, and the difference is statistically significant at the .05 level. On further inspection, however, table 1.20 provides less clear cut evidence that the Americans' responses exhibit higher life satisfaction. American respondents are overrepresented in both extremes, in both the very satisfied and the unsatisfied categories. If the top two categories on the satisfaction scale (very satisfied and satisfied) are combined, the French actually indicate higher life satisfaction: 83 percent versus 77 percent. Thus, it is unclear from these data whether the French are less satisfied or less prone to use the extreme ends of the scales. The propensity to express oneself in extremes can be influenced by cultural and social expectations. Cultural and social norms may discourage French women from reporting themselves as very satisfied compared with Americans. 1.8.3 Comparing SWB with the U-Index The U-index is less susceptible to a tendency for the Americans to be more emphatic than the French as long as both apply their interpretation of the scales consistently to positive and negative emotions. To take an extreme example, suppose the French only use the zero to five portion of the zero to six scale, while the Americans utilize the full scale. Provided that the French 38. See http://management.ucsd.edu/faculty/directory/schkade/fa-study/.
72 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.20
Distribution of reported life satisfaction in Columbus, OH and Rennes, France
U.S. (%)
France (%)
Not at all satisfied
1.6
1.1
Not very satisfied
21.4
16.1
Satisfied
51.0
70.0
Very satisfied
26.1
12.9
Notes: Life satisfaction is based on the question, "Taking all things together, how satisfied are you with your life as a whole these days?" Sample size is 810 women for Columbus and 816 women for Rennes. Chi-square test of identical distributions rejects at p 0.001.
use the zero to five range consistently for reporting positive and negative emotions--that is, an emotion reported as a five is always experienced more intensively than an emotion reported as a four--then, apart from integer concerns, the U-index is unaffected by this differential use of scales. As commonly applied, however, the standard life satisfaction measure is not robust to such reporting differences across people because the French would appear as less satisfied if they express themselves less emphatically. The first row of table 1.21 reports the average episode-level U-index for the two samples. In this case, the U-index for an episode is defined as equal to one if the maximum rating of "tense/stressed," "depressed/blue," or "irritated/angry" strictly exceed the rating of "happy," and zero if not. The U-index was weighted by the proportion of each person's waking day spent in an episode to derive an overall estimate. In contrast to reported life satisfaction, the U-index is 2.8 percentage points lower in the French sample (16 percent) than in the American sample (18.8 percent). Thus, the French appear to spend less of their time engaged in unpleasant activities (i.e., activities in which the dominant feeling is a negative one) than do the Americans in our samples. We explored whether the lower U-index for the French is a result of any single negative emotion, or combinations of them. The lower U-index for the French appears to be a fairly robust result. If we required that at least two negative feelings were rated more strongly than happy, for example, the U-index was still 2.8 points lower in France than in the United States (10.1 percent versus 7.4 percent) And if we dropped any one of the negative emotions and compared the remaining two to happy, the U-index was lower in France than in the United States in each case. These results suggest that the lower U-index in France is not due to the rating of any particular negative emotion in our study. The other rows of table 1.21 provide comparisons of the episode-level U-index for various subpopulations. The general pattern is sensible. For example, the U-index in both countries is considerably lower on weekends
National Time Accounting: The Currency of Life 73
Table 1.21
U-index for various groups in Columbus, OH and Rennes, France DRM surveys
Group All Enrollment status Nonstudent Student Employment status Employed Unemployed Household income Bottom half Top half Day of week Weekday Weekend
U.S. 0.188 0.181 0.243 0.189 0.219 0.203 0.169 0.205 0.144
France 0.160 0.144 0.229 0.143 0.190 0.173 0.143 0.174 0.122
Difference 0.028 0.037 0.014 0.046 0.029 0.030 0.026 0.031 0.022
Notes: U-index is computed as proportion of time in which the rating of the maximum of tense, blue, and angry is strictly greater than the rating of happy. P-values are for test of country differences for each group. Significant at the 1 percent level. Significant at the 5 percent level. Significant at the 10 percent level.
than on weekdays. The French-American gap is largest for nonstudents, employed people, low-income people, and during the week. Interestingly, in both countries--but especially in the United States--the U-index of the unemployed is much higher during the week than it is during weekends. This pattern suggests that observing others go to work during the week worsens the mood of the unemployed during weekdays. There is greater inequality in the U-index across people in the American sample than in the French sample. Figure 1.12 displays the average U-index by quintile of the individual-level U-index distribution in each country. The average woman in Columbus in the top quintile of the distribution spent 57.5 percent of her time in an unpleasant state, while her counterpart in Rennes spent 49.0 percent of her time in an unpleasant state. Regression analysis indicated that the gap in the upper tail is only partially accounted for by independent variables such as the log of household income, a quadratic in age, school enrollment, and day of week. Controlling for these variables reduced the U.S.-French gap in the upper quintile from 8.5 points to 5.3 points. Another issue concerns vacations. In our sample, the French report taking twenty-one more vacation days than the Americans. We were not able to interview people if they were away from home, so we did not sample most vacation days. Accounting for vacations would almost certainly lower the U-index in France relative to the United States, as vacation days are likely to have a lower U-index than nonvacation days. The following back of the
74 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Fig. 1.12 Average U-index by quintile of the U-index distribution in U.S. and France based on DRM surveys envelope calculation suggests, however, that this is not a large bias. The twenty-one day difference in vacations amounts to only 5.8 percent of the year. If the U-index is 10 points lower on vacation days than nonvacation days, which is almost double the difference on weekdays and weekends, then the French U-index would be an additional 0.58 percentage points lower than the American U-index. 1.8.4 Counterfactual Cross-Country Comparisons: Activity Level Analysis Table 1.22 presents the U-index for twenty-one activities and the proportion of the day the average person devoted to each activity based on the DRM. (These activities are different from those in some of our other DRMs because of translation issues.) If more than one activity was engaged in at a time, we selected the activity that was indicated by respondents as being most important at the time. Activities such as working, commuting, and child care have a high U-index, and activities such as walking, making love, and exercising have a low U-index, similar to our earlier findings. Both the pattern of time allocation and the U-index for each activity are similar in the two countries, with correlations of 0.93 and 0.85, respectively. The most notable exceptions to this pattern are that the Americans find child care substantially more unpleasant than do the French, and the French spend less time engaged in child care and more time eating. The latter is explained mainly by the fact that Americans are much less likely to indicate eating as their main activity when they engage in multiple activities that include eating. It is also worth noting that the French women in our sample are slightly less likely to have children living at home (56 percent versus 60 percent).
National Time Accounting: The Currency of Life 75
Table 1.22
The U-index and allocation of time across activities based on DRM surveys
U-index per activity
Percent of time (%)
Focal activity
U.S.
France
U.S.
France
Walking
0.04
0.09
0.63
1.69
Making love
0.05
0.03
0.77
0.98
Exercise
0.06
0.03
0.88
1.21
Playing
0.07
0.02
1.47
1.26
Reading, nonwork
0.09
0.07
2.97
4.36
Eating
0.10
0.09
5.22
11.11
Prayer
0.11
0.16
1.70
0.25
TV
0.12
0.14
7.07
7.32
Relaxing
0.13
0.13
2.88
2.85
Preparing food
0.14
0.13
2.92
3.29
Talking, nonwork
0.14
0.12
9.35
11.58
Grooming
0.15
0.14
5.19
4.76
Other
0.16
0.13
8.54
5.72
Housework
0.18
0.23
5.91
5.16
Sleep
0.18
0.15
2.70
2.32
Other travel
0.20
0.20
3.23
3.22
Shop
0.22
0.20
4.86
4.35
Computer, nonwork
0.23
0.22
2.52
2.28
Child care
0.24
0.11
6.85
4.50
Commute
0.27
0.26
2.22
1.68
Work
0.29
0.26
22.10
20.12
The data in table 1.22 can be used to perform counterfactual calculations. Specifically, we can use the time allocation across activities for one country to weight the U-index for the other country and thus create a "synthetic" U-index. To be more precise, define the synthetic U-index using country j's time allocation (Hji) and country k's U-index (Uik) for activities denoted i as Uj,k i HjiUik. The "synthetic" U-index indicates how the average French woman, say, would feel if she experienced her activities in the same way as the average American woman. Table 1.23 reports the synthetic U-indexes for each country.39 The results indicate that if the French and American women's allocation of time is weighted by either the average American woman's rating of activities or the average French woman's rating of activities, the average French woman is predicted to have a lower synthetic U-index than the average American woman. But only about one-third to 40 percent of the between- 39. Notice that when the same country's time allocation and activity-level U-indexes are used the synthetic U-index is slightly different from the episode-level U-indexes reported in the first row of table 1.8. This discrepancy arises because there is a weak correlation between time allocation and the U-index at the individual level.
76 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone
Table 1.23
Synthetic U-index based on country's aggregate time allocation and country's U-index by activity
Country's time
Country's U-index
U.S.
France
Difference
t-ratio
U.S. France
0.189
0.177
0.012
1.02
0.169
0.159
0.010
0.90
Notes: Standard errors for t-ratios are derived from a bootstrap procedure that takes into account sampling variability in the U-index and in the time allocation. Calculations based on data in table 1.3.
country difference in the U-index comes about because of differences in time allocation. Moreover, with small samples to compute time allocation, the difference in the synthetic U-index is not statistically significant regardless of which country's activity ratings are used. We can calculate the synthetic U-indexes using larger samples of time allocation data from national time-use surveys, however. This provides a check on whether our results for Rennes and Columbus can be extended to the countries as a whole, and yields more precise estimates. Specifically, we analyzed national time-use data on American women from the 2003 to 2004 ATUS and on French women from the 1998 to 1999 Enquкte Emploi du Temps survey by INSEE. We restrict both samples to women age eighteen to sixty. Although the French data are from an earlier time period, they are the most recent national data publicly available, and time allocation does not change very rapidly over time within countries. Because the activity categories in national time-use data are not harmonized, we collapsed the activities in these surveys into six broad categories: work, compulsory activities, active leisure, passive leisure, eating, and other. The U-index for these categories was computed from the DRM for Rennes and Columbus for the same activities. Results are reported in table 1.24. The national time allocations are generally similar to what we found for Rennes and Columbus. In particular, using national data the French women spend less time working, less time participating in passive leisure (e.g., watching TV), and more time participating in active leisure (e.g., exercise and reading) and eating than do the American women. As was found before, the French allocation of time produces a slightly lower synthetic U-index regardless of whether the American or the French U-index is used to rate each activity. Using either U-index to rate the activities, the French allocation of time produces about a 1 percentage point lower synthetic U-index. With the larger national time-use samples, the differences are statistically significant at the 0.10 level, although they are similar in magnitude to the differences reported in table 1.23.
National Time Accounting: The Currency of Life 77
Table 1.24
National time-use data for U.S. and France and synthetic U-indices
Work/commute Compulsory Passive
Active Eating Other
(%)
(%)
leisure (%) leisure (%) (%)
(%)
U.S. France U.S. France
Fraction of awake time spent in each activity
24.6
35.2
24.8
7.5
21.8
34.8
18.1
10.6
Average U-index per activity
0.29
0.19
0.15
0.10
0.26
0.17
0.14
0.09
6.6
1.3
14.3
0.5
0.10 0.15 0.09 0.13
Notes: Synthetic U-index based on country's aggregate time allocation from national time-use data and country's U-index by activity from DRM. Country's time
Country's U-index
U.S.
France
Difference
t-ratio
U.S. France
0.193
0.184
0.010
1.67
0.173
0.164
0.009
1.74
Standard errors for t-ratios are derived from a bootstrap procedure that takes into account sampling variability in the U-index and in the time allocation. The work activity combines working and commuting; the compulsory activity combines shopping, housework, preparing food, and grooming; passive leisure combines watching TV, nonwork computer use, relaxing, and napping; activity leisure combines exercise, walking, making love, playing, and talking.
1.9 Conclusion National Time Accounting provides a method for tracking time allocation and assessing whether people are experiencing their daily lives in more or less enjoyable ways. This chapter demonstrates how NTA can be used to compare groups of individuals, countries and eras. Many economists argue that a decline in the amount of time spent working has been a major source of improvement in Americans' daily lives over the last century (Fogel 1999). Shifts in time use among nonwork activities also affect the experience of daily life. If nonwork time increases in the next century as much as it did in the last century, it will be even more important to understand the experience of nonwork time. Tracking the U-index over time, either at the episode level or at the activity level, provides a means for measuring whether daily life is becoming more or less pleasant, and of understanding why. To facilitate NTA in the future, we think that adding a module on affective experience to ongoing time-use surveys, such as ATUS, should be a priority. The PATS data on evaluated time use that we developed for NTA and summarize here reinforce some findings from the previous literature on overall happiness and life satisfaction and provide new results and puzzles. At the individual level within a country, the demographic correlates of experi-
78 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone enced well-being and life (or happiness) satisfaction mostly have the same sign. Life satisfaction and the U-index, however, yield a different ranking of France and the United States, most likely because of cultural differences in reporting that lead the French to appear less satisfied. In addition, experienced well-being measures provide a means for decomposing differences between groups that is not possible with conventional life satisfaction data. For example, we show how differences in subjective well-being between age groups can be attributed to a component due to differences in time allocation and a component due to differences in feelings for a given set of activities. This analysis revealed that differences in time use account for a majority of the difference in experienced well-being between younger and older individuals. Unlike previous attempts to measure experienced wellbeing in the time-use literature, we emphasize that subjective-well being is multidimensional, and propose the U-index as a simple means to reflect the nonlinear relationship among emotions in a National Time Accounting framework. Like the NIPAs, NTA is a descriptive, not prescriptive, technique. The method of NTA does not lead to immediate policy recommendations. For example, the fact that spending time socializing may be more enjoyable than working for pay for the average person does not necessarily lead to the recommendation that people should socialize more and work less. Paid work is obviously required to afford a certain lifestyle. A similar limitation applies to the NIPAs: although national income would be increased if all workers trained for higher paying professions, there are psychic and monetary costs that must be taken into account before making such a policy recommendation. To draw policy conclusions, we would recommend using the PATS or related instruments to measure outcomes of policy relevant experiments, such as the Moving to Opportunities public housing experiment. Existing time-use data sets provide several opportunities for additional applications of NTA. One possibility is to use the harmonized international time-use data sets to compare how people in different countries devote time to various activities and to evaluate the activities by their average emotional experience according to the PATS. The clusters of activities identified in section 1.6 would seem particularly appropriate for comparing time use across countries. Another possibility is to use existing time-use data for the United States to study the effect of aging on the allocation of time across activities by following cohorts as they age. Again, the clusters of similar activities identified in section 1.6 could facilitate the analysis. Several extensions, unresolved issues, and research issues concerning NTA should also be noted. First, although we based the emotions that we surveyed partly on the Russell circumplex and partly on practicality, the precise set of emotions could be tailored for the particular application at hand. For example, studies related to health and aging might focus on feel-
National Time Accounting: The Currency of Life 79 ings of aches, pain, weariness, fatigue, and disorientation. In addition, PATS might be adapted to measure people's sense of purpose about their daily routines. People could be asked whether they considered their use of time during sampled episodes to be meaningful or a waste of time. If additional emotions are included, the robustness of the U-index to the set of surveyed emotions can be further explored, although some features of experience (e.g., meaningfulness) would seem to represent separate subjective components of well-being. Another issue concerns the context of time use. That is, the precise situations that people are engaged in during their daily activities. Available timeuse surveys collect only coarse information on the nature of activities. The fact that activity dummies account for such a small share of the variability in affective experience suggests that important features of activities are not measured by time-use surveys. Thus, tracking the change in activities over time weighted by the activity-level U-index (or some other activitylevel measure of emotional experience) is susceptible to missing important changes in people's affective experiences because a great deal of what generates emotional experience occurs within a given set of measured activities. A related issue is that the nature of some activities changes over time. For example, the experience of television viewing is likely to be quite different today than forty years ago, when there were few channels, television sets were black and white, and Tivo was not available to skip over commercials. While changes in the nature of activities present a problem for all studies that track time use over historical time, the problems are particularly apparent for NTA. In some respects, the problem is akin to changes in product quality in the consumer price index. The prospect of tracking affective experience at the episode-level in the future, however, provides a way to avoid problems caused by changes in the nature of activities because it would not depend on the a priori assignment of activities. In addition, a time-series of episodelevel data on affective experience would enable research into the changing hedonic nature of activities. Data on emotional experience might also be used to explain people's choices. What types of preferences are consistent with observed time allocation patterns if people seek to maximize some function of their flow of emotional experiences? What other considerations besides maximization of emotional experience is needed to rationalize observed choices about time allocation in a maximizing framework? Or, if maximization is considered too strong an assumption, can people's time allocation be explained by a small set of heuristics? Of course, modeling behavior with data on subjective well-being requires that information on a relevant set of emotional experiences is collected. It should also be noted that understanding people's choices is not a prerequisite for NTA, just as understanding choices about work, consumption, and investment are not a prerequisite for the NIPAs.
80 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Nonetheless, the evaluated time-use data provide a new opportunity to model people's allocation of time. Finally, it is unclear how to fully integrate sleep and health into NTA. To some extent, both factors are reflected in our measures of affect. For example, people who are in poor health experience more pain during their daily lives (Krueger and Stone 2008). And a bad night sleep is associated with a bad mood and greater tiredness throughout the day (Kahneman et al. 2004). In other words, sleep and health both affect the process benefit of various uses of time. But if people learn to sleep half as much without lowering their average emotional experience during waking moments, our current summary measures would not credit an improvement in well-being. In addition, health surely has a direct effect on well-being independent of any effect on momentary emotional experience. While these limitations of NTA are important, they are not insurmountable. We suspect that many of the current limitations of NTA are amenable to research, just as research helped to overcome some of the problems posed by changes in product quality in the NIPAs. Moreover, the choices that people make regarding their allocation of time, particularly labor supply, have long been subject to economic analysis. Research on the allocation and experience of nonwork time is less developed, but no less important for economics and policy. Evaluated time use also strikes us as a fertile area for research because most determinants of subjective well-being are not well captured by data on market transactions, and this will be even more so in the future as people live longer and spend a smaller share of their lives engaged in market work and home production.
National Time Accounting: The Currency of Life 81
Appendix
Linear probability multiple regression models for U-index, full sample, and by sex
Full sample
Women
Men
Explanatory variable Coefficient t-ratio Coefficient t-ratio Coefficient t-ratio
Female Black Hispanic Log income High school Some college College College Age Age-squared Married Tuesday Wednesday Thursday Friday Saturday Sunday June July August No. of episodes
0.024 0.052 0.033 ­0.023 ­0.005 ­0.017 ­0.056 ­0.045 0.003 0.000 ­0.017 ­0.012 0.004 0.005 ­0.020 ­0.009 ­0.061 ­0.015 ­0.025 0.046 9,989
1.96 1.84 1.26 ­2.31 ­0.21 ­0.96 ­3.27 ­2.31 1.67 ­1.95 ­1.17 ­0.51 0.18 0.22 ­0.86 ­0.36 ­2.62 ­0.92 ­1.67 2.32
-- 0.042 0.057 ­0.027 0.006 0.006 ­0.045 ­0.020 0.009 0.000 ­0.051 0.019 0.026 0.035 0.000 0.027 ­0.052 ­0.036 ­0.022 0.030 6,136
-- 1.18 1.62 ­2.15 0.17 0.27 ­1.99 ­0.71 3.11 ­3.58 ­2.69 0.61 0.89 1.13 0.00 0.82 ­1.79 ­1.66 ­1.01 1.16
0.065 0.010 ­0.020 ­0.010 ­0.052 ­0.070 ­0.082 0.000 0.000 0.020 ­0.043 ­0.022 ­0.024 ­0.049 ­0.055 ­0.070 0.010 ­0.031 0.065 3,853
1.42 0.25 ­1.21 ­0.27 ­1.98 ­2.76 ­3.00 ­0.02 ­0.07 0.94 ­1.28 ­0.62 ­0.68 ­1.42 ­1.52 ­1.87 0.41 ­1.50 2.14
Notes: All regressions also control for 15 "who with" dummies, 5 dummies indicating the order in which affect questions were asked, and an intercept. Heteroskedasticity consistency standard errors that allow for within-person correlated errors were calculated. Data are from the PATS. Dashed cells indicate there is no coefficient, since the gender variable is a constant for women and men.
References Aguiar, M., and E. Hurst. 2007. Measuring trends in leisure: The allocation of time over five decades. Quarterly Journal of Economics 122 (3): 969­1006. Alesina, A. F., E. L. Glaeser, and B. Sacerdote. 2002. Why doesn't the U.S. have a European-style welfare state? Brookings Papers on Economic Activity (Fall): 187­ 277. Ashby, F. G., A. M. Isen, and A. U. Turken. 1999. A neuropsychological theory of positive affect and its influence on cognition. Psychological Review 106 (3): 529­50. Becker, G. 1965. A theory of the allocation of time. Economic Journal 75 (299): 493­517. Berk, L. S., D. L. Felten, S. A. Tan, B. B. Bittman, and J. Westengard. 2001. Modulation of neuroimmune parameters during the eustress of humor-associated mirthful laughter. Alternative Therapies 7 (2): 62­76. Berkowitz, L. 1987. Mood, self-awareness, and willingness to help. Journal of Personality and Social Psychology 52 (4): 721­29.
82 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Blair, E., and S. Burton. 1987. Cognitive processes used by survey respondents to answer behavioral frequency questions. Journal of Consumer Research 14 (2): 280­88. Blanchflower, D. G. 2007. Is unemployment more costly than inflation? NBER Working Paper no. W13505. Cambridge, MA: National Bureau of Economic Research, October. Brickman, P., D. Coates, and R. Janoff-Bulman. 1978. Lottery winners and accident victims: Is happiness relative? Journal of Personality and Social Psychology 36 (8): 917­27. Clark, A., E. Diener, Y. Georgellis, and R. Lucas. 2003. Lags and leads in life satisfaction: A test of the baseline hypothesis. Working Paper 2003-14, DELTA (Ecole normale supйrieure). Clark, A., and A. Oswald. 1994. Unhappiness and unemployment. The Economic Journal 104 (424): 648­59. Coghill, R. C., J. G. McHaffie, and Y. Yen. 2003. Neural correlates of interindividual differences in the subjective experience of pain. Proceedings of the National Academy of Sciences 100 (14): 8538­542. Connolly, M. 2007. Some like it mild and not too wet: The influence of weather on subjective well-being. In It's about time: Three essays on time use, weather and well-being, PhD diss., Department of Economics, Princeton University. Csikszentmihalyi, M. 1990. Flow: The psychology of optimal experience. New York: HarperCollins. Csikszentmihalyi, M., J. D. Patton, and M. Lucas. 1997. Le bonheur, l'experience optimale et les valeurs spirituelles: Une йtude empirique auprиs d'adolescents (Happiness, the optimal experience, and spiritual values: An empirical study of adolescents). Revue Quebecoise de Psychologie 18: 167­90. Cunningham, M. R., J. Steinberg, and R. Grev. 1980. Wanting to and having to help: Separate motivations for positive mood and guilt-induced helping. Journal of Personality and Social Psychology 38 (2): 181­92. Danner, D. D., D. A. Snowdon, and W. V. Friesen. 2001. Positive emotions in early life and longevity: Findings from the nun study. Journal of Personality and Social Psychology 80 (5): 804­13. Devins, G. M., J. Mann, H. P. Mandin, and C. Leonard. 1990. Psychosocial predictors of survival in end-stage renal disease. Journal of Nervous and Mental Disease 178 (2): 127­33. Diener, E., and R. E. Lucas. 1999. Personality and subjective well-being. In Wellbeing: The foundations of hedonic psychology, ed. D. Kahneman, E. Diener, and N. Schwarz, 213­29. New York: Russell-Sage. Diener E., R. E. Lucas, and C. N. Scollon. 2006. Beyond the hedonic treadmill: Revisions to the adaptation theory of well-being. American Psychologist 61: 305­14. Diener, E., C. Nickerson, R. E. Lucas, and E. Sandvik. 2002. Dispositional affect and job outcomes. Social Indicators Research 59 (3): 229­59. Diener, E., E. Sandvik, and W. Pavot. 1991. Happiness is the frequency, not the intensity, of positive versus negative affect. In Subjective well-being, ed. F. Strack, M. Argyle, and N. Schwarz, 119­40. New York: Pergamon Press. Dillon, K. M., B. Minchoff, and K. H. Baker. 1985. Positive emotional states and enhancement of the immune system. International Journal of Psychiatry in Medicine 15 (1): 13­18. Dow, G. K., and F. T. Juster. 1985. Goods, time, and well-being: The joint dependence problem. In Time, Goods, and Well-Being, ed. F. T. Juster and F. P. Stafford. Survey Research Center, Institute for Social Research 397­413. Ann Arbor, MI: University of Michigan Press. Doyle, W. J., D. A. Gentile, and S. Cohen. 2006. Emotional style, nasal cytokines,
National Time Accounting: The Currency of Life 83 and illness expression after experimental rhinovirus exposure. Brain, Behavior, and Immunity 20 (2): 175­81. Ekman, P., and E. L. Rosenberg, eds. 1997. What the face reveals. New York: Oxford University Press. Evans, P., M. Bristow, F. Hucklebridge, A. Clow, and N. Walters. 1993. The relationship between secretory immunity, mood and life-events. British Journal of Clinical Psychology 32 (2): 227­36. Fogel, R. W. 1999. Catching up with the economy. American Economic Review 89 (1): 1­21. ------. 2001. Simon S. Kuznets, Biographical memoirs, vol. 79. Washington DC: The National Academy Press. Frey, B., and A. Stutzer. 2002. Happiness and economics. Princeton, NJ: Princeton University Press. Gershuny, J., and B. Halpin. 1996. Time use, quality of life and process benefits. In In pursuit of the quality of life, ed. 188­210. Avner Offer, Oxford: Clarendon Press. Glorieux, I. 1993. Social interaction and the social meanings of action: A time budget approach. Social Indicators Research 30 (2­3): 149­73. Graham, C., A. Eggers, and S. Sukhtankar. 2006. Does happiness pay? An exploration based on panel data from Russia. Journal of Economic Behavior and Organization 55 (3): 319­42. Gross, J. J., L. L. Carstensen, M. Pasupathi, J. Tasi, C. G. Skorpen, and A. Y. Hsu. 1997. Emotion and aging: Experience, expression and control. Psychol Aging 12 (4): 590­99. Harker, L., and D. Keltner. 2001. Expressions of positive emotions in women's college yearbook pictures and their relationship to personality and life outcomes across adulthood. Journal of Personality and Social Psychology 80: 112­24. Heady, B. W., and A. J. Wearing. 1989. Personality, life events, and subjective wellbeing: Toward a dynamic equilibrium model. Journal of Personality and Social Psychology 57 (4): 731­39. Howell, R. T., M. L. Kern, and S. Lyubomirsky. 2007. Health benefits: Metaanalytically determining the impact of well-being on objective health indicators. Health Psychology Review 1: 83­136. Isen, A. M., and P. F. Levin. 1972. Effect of feeling good on helping: Cookies and kindness. Journal of Personality and Social Psychology 21: 384­88. Juster, F. T. 1985. Preferences for work and leisure. In Time, goods, and well-being. ed. F. T. Juster and F. P. Stafford, 333­51. Ann Arbor, MI: Institute for Social Research, University of Michigan. Juster, F. T., P. Courant, and G. K. Dow. 1985. A conceptual framework for the analysis of time allocation data. In Time, goods, and well-being, ed. F. T. Juster and F. P. Stafford. Ann Arbor, MI: Survey Research Center, Institute for Social Research, University of Michigan. Juster, F. T., and F. P. Stafford, eds. 1985. Time, goods, and well-being. Ann Arbor: Institute for Social Research. Kahneman, D. 2003. Maps of bounded rationality: Psychology for behavioral economics. American Economic Review 93 (5): 1449­75. Kahneman, D., and A. Krueger. 2006. Developments in the measurement of subjective well-being. Journal of Economic Perspectives 20 (21): 3­24. Kahneman, D., A. Krueger, D. Schkade, N. Schwarz, and A. Stone. 2004. A survey method for characterizing daily life experience: the Day Reconstruction Method. Science 306 (December): 1776­80. Kahneman, D., P. Wakker, and R. Sarin. 1997. Back to Bentham? Explorations of experienced utility. Quarterly Journal of Economics 112 (2): 375­405.
84 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Kiecolt-Glaser, J., L. D. Fisher, P. Ogrocki, J. C. Stout, C. E. Speicher, and R. Glaser. 1988. Marital quality, marital disruption, and immune function. Psychosomatic Medicine 49 (1): 13­34. Kirschbaum, C., K. M. Pirke, and D. H. Hellhammer. 1993. The "Trier Social Stress Test"--A tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology 28 (1­2): 76­81. Krueger, A. B. 2007. Are we having more fun yet? Categorizing and evaluating changes in time allocation. Brookings Papers on Economic Activity 2 (September): 193­215. Krueger, A. B., D. Kahneman, C. Fischler, D. Schkade, N. Schwarz, and A. Stone. 2009. Comparing time use and subjective well-being in France and the U.S. Social Indicators Research, forthcoming. Krueger, A., and A. Mueller. 2008. The lot of the unemployed: A time use perspective. IZA Discussion Paper no. 3490, Institute for the Study of Labor (IZA), Bonn, Germany. Krueger, A. B., and D. Schkade. 2008. The reliability of subjective well-being measures. Journal of Public Economics 92 (8­9): 1833­45. Krueger, A. B., and A. Stone. 2008. Assessment of pain: A community-based diary survey. The Lancet 371 (9623): 1510­25. Kubey, R., and M. Csikszentmihalyi. 1990. Television and the quality of life: How viewing shapes everyday experience. Hillsdale, NJ: Lawrence Erlbaum. Layard, R. 2005. Happiness: Lessons from a new science. London: Penguin. Leonhardt, D. Obamanomics. New York Times Magazine, August 20, 2008. Avail- able at: http://www.nytimes.com/2008/08/24/magazine/24Obamanomics-t.html ?pagewantedall. Levy, S. M., J. Lee, C. Bagley, and M. Lippman. 1988. Survival hazard analysis in first recurrent breast cancer patients: Seven-year follow-up. Psychosomatic Medicine 50: 520­28. Lucas, R. E. 2001. Pleasant affect and sociability: Towards a comprehensive model of extraverted feelings and behaviors. Dissertation Abstracts International 61 (10B): 5610 (UMI no. AAI9990068.) Lucas, R. E., A. E. Clark, Y. Georgellis, and E. Diener. 2003. Reexamining adaptation and the set point model of happiness: Reactions to changes in marital status. Journal of Personality and Social Psychology 84 (3): 527­39. Lykken, D. 1999. Happiness: What studies on twins show us about nature, nurture and the happiness set-point. New York: Golden Books. Lyubomirsky, S., L. King, and E. Diener. 2005. The benefits of frequent positive affect: Does happiness lead to success? Psychological Bulletin 131 (6): 803­55. Lyubomirsky, S., K. M. Sheldon, and D. Schkade. 2005. pursuing happiness: The architecture of sustainable change. Review of General Psychology 9 (2): 111­31. Maier, H., and J. Smith. 1999. Psychological predictors of mortality in old age. Journal of Gerontology 54B: 44­54. March, J. 1978. Bounded rationality, ambiguity, and the engineering of choice. Bell Journal of Economics 9 (2): 587­608. Marks, G. N., and N. Fleming. 1999. Influences and consequences of wellbeing among Australian young people: 1980­1995. Social Indicators Research 46: 301­23. Marsland, A. L., S. Pressman, and S. Cohen. 2007. Positive affect and immune function. In Psychoneuroimmunology, 4th ed., vol. 2, ed. R. Ader, 761­779. San Diego, CA: Elsevier. Michelson, W. 2005. Time use: Expanding explanation in the social sciences. Boulder, CO: Paradigm Publishers.
National Time Accounting: The Currency of Life 85 Nadal, J. I. G., and A. S. Sanz. 2007. A note on leisure inequality in the U.S.: 1965­ 2003. Oxford University, Department of Economics, Discussion Paper. O'Malley, M. N., and L. Andrews. 1983. The effect of mood and incentives on helping: Are there some things money can't buy? Motivation and Emotion 7 (2): 179­89. Oswald, A., and N. Powdthavee. Does happiness adapt? A longitudinal study of disability with implications for economists and judges. IZA Discussion Paper no. 2208, Institute for the Study of Labor (IZA), Bonn, Germany. Palmore, E. B. 1969. Predicting longevity: A follow-up controlling for age. Gerontologist 9 (4): 247­50. Perera, S., E. Sabin, P. Nelson, and D. Lowe. 1998. Increases in salivary lysozyme and IgA concentrations and secretory rates independent of salivary flow rates following viewing of humorous videotape. International Journal of Behavioral Medicine 5 (2): 118­28. Polk, D. E., S. Cohen, W. J. Doyle, D. P. Skoner, and C. Kirschbaum. 2005. State and trait affect as predictors of salivary cortisol in healthy adults. Psychoneuroendocrinology 30 (3): 261­72. Pollak, R., and M. L. Wachter. 1975. The relevance of the household production function and its implications for the allocation of time. Journal of Political Economy 83 (2): 255­77. Ramey, V., and N. Francis. 2006. A century of work and leisure. NBER Working Paper no. 12264. Cambridge, MA: National Bureau of Economic Research, May. Riis, J., G. Loewenstein, J. Baron, C. Jepson, A. Fagerlin, and P. A. Ubel. 2005. Ignorance of hedonic adaptation to hemo-dialysis: A study using ecological momentary assessment. Journal of Experimental Psychology: General 134 (1): 3­9. Robinson, J., and G. Godbey. 1997. Time for life: The surprising ways Americans use their time. University Park, PA: Pennsylvania State University Press. Robinson, M. D., and G. L. Clore. 2002. Belief and feeling: Evidence for an accessibility model of emotional self-report. Psychological Bulletin 128: 934­60. Russell, J. A. 1980. A circumplex model of affect. Journal of Personality and Social Psychology 39: 1161­78. Schwarz, N. 2007. Retrospective and concurrent self-reports: The rationale for realtime data capture. In The science of real-time data capture: Self-reports in health research, ed. A. A. Stone, S. S. Shiffman, A. Atienza, and L. Nebeling, 11­26. New York: Oxford University Press. Schwarz, N., and G. L. Clore. 1983. Mood, misattribution, and judgments of wellbeing: Informative and directive functions of affective states. Journal of Personality and Social Psychology 45 (3): 513­23. Schwarz, N., D. Kahneman, and J. Xu. 2009. Global and episodic reports of hedonic experience. In Using calendar and diary methods in life events research, ed. R. Belli, D. Alwin, and F. Stafford, 157­74. Newbury Park, CA: SAGE. Schwarz, N., and F. Strack. 1999. Reports of subjective well-being: Judgmental processes and their methodological implications. In Well-being: The foundations of hedonic psychology, ed. D. Kahneman, E. Diener, and N. Schwarz, 61­84. New York: Russell-Sage. Shapiro, D., L. D. Jamner, and I. B. Goldstein. 1997. Daily mood state and ambulatory blood pressure. Psychophysiology 34: 399­405. Smith, D. M., K. M. Langa, M. U. Kabeto, and P. A. Ubel. 2005. Health, wealth, and happiness. Psychological Science 16 (9): 663­66. Smyth, J., M. C. Ockenfels, L. Porter, C. Kirschbaum, D.C. Hellhammer, and A. A. Stone. 1998. Stressors and mood measured on a momentary basis are associated with salivary cortisol secretion. Psychoneuroendocrinology 23: 353­70.
86 A. B. Krueger, D. Kahneman, D. Schkade, N. Schwarz, and A. A. Stone Snowdon, D. 2001. Aging with grace. New York: Bentam. Spanier, G. B., and F. F. Furstenberg. 1982. Remarriage after divorce: A longitudinal analysis of well-being. Journal of Marriage and the Family 44 (3): 709­20. Stone, A. A. 1987. Event content in a daily survey differentially predicts mood. Jour- nal of Personality and Social Psychology 52: 56­58. Stone, A. A., D. S. Cox, H. Valdimarsdottir, L. Jandorf, and J. M. Neale. 1987. Evi- dence that secretory IgA antibody is associated with daily mood. Journal of Personality and Social Psychology 52 (5): 988­93. Stone, A. A., S. M. Hedges, J. M. Neale, and M. S. Satin. 1985. Prospective and cross-sectional mood reports offer no evidence of a "blue Monday" phenomenon. Journal of Personality and Social Psychology 49: 129­34. Stone, A. A., C. Marco, C. E. Cruise, D. S. Cox, and J. M. Neale. 1996. Are stressinduced immunological changes mediated by mood? A closer look at how both desirable and undesirable daily events influence sIgA antibody. International Journal of Behavioral Medicine 3 (1): 1­13. Stone, A. A., J. E. Schwartz, N. Schwarz, D. Schkade, A. Krueger, and D. Kahneman. 2006. A population approach to the study of emotion: Diurnal rhythms of a working day examined with the Day Reconstruction Method (DRM). Emotion 6: 139­49. Stone, A. A., and S. Shiffman. 1994. Ecological Momentary Assessment (EMA) in behavioral medicine. Annals of Behavioral Medicine 16 (3): 199­202. Stone, A. A., S. Shiffman, A. Atienza, and L. Nebeling. eds. 2007. The science of real-time data capture: Self-reports in Health Research. London: Oxford University Press. Stone, A. A., S. S. Shiffman, and M. W. DeVries. 1999. In Well-being: The foundations of hedonic psychology, ed. D. Kahneman, E. Diener, and N. Schwarz, 26­39. New York: Russell-Sage. Stone, A. A., H. B. Valdimarsdottir, E. S. Katkin, J. M. Burns, D. S. Cox, S. Lee, J. Fine, D. Ingle, and D. H. Bovbjerg. 1993. Mitogen-induced lymphocyte responses are reduced following mental stressors in the laboratory. Psychology and Health 8: 269­84. Sudman, S., and B. Wansink. 2002. Consumer panels. Chicago: American Marketing Association. Ubel, P. A., G. Loewenstein, N. Schwarz, and D. Smith. 2005. Misimagining the unimaginable: The disability paradox and health care decision making. Health Psychology 24 (4, Supplement): S57­S62. Urry, H. L., J. B. Nitschke, I. Dolski, D.C. Jackson, K. M. Dalton, C. J. Mueller, M. A. Rosenkranz, C. D. Ryff, B. H. Singer, and R. J. Davidson. 2004. Making a life worth living: Neural correlates of well-being. Psychological Science 15 (6): 367­72. Vitaliano, P. P., J. M. Scanlan, H. D. Ochs, K. Syrjala, I. C. Siegler, and E. A. Snyder. 1998. Psychological stress moderates the relationship of cancer history with natural killer cell activity. Annals of Behavioral Medicine 20 (3): 199­208. Wong, G. 2007. Be it ever so humble: Understanding housing using subjective wellbeing data. University of Pennsylvania, Wharton School, Working Paper. Xu, J., and N. Schwarz. 2009. Do we really need a reason to indulge? Journal of Marketing Research 46 (1): 25­36. Zuckerman, D. M., S. V. Kasl, and A. M. Ostfeld. 1984. Psychosocial predictors of mortality among the elderly poor. The role of religion, well-being, and social contacts. American Journal of Epidemiology 119 (3): 410­23.

File: national-time-accounting-the-currency-of-life.pdf
Title: Measuring the Subjective Well-Being of Nations: National Accounts of Time Use and Well-Being
Author: Alan B. Krueger (Editor)
Published: Fri May 20 11:20:22 2011
Pages: 79
File size: 1.38 Mb


A history of bestiality, 140 pages, 1.01 Mb

García Márquez, Gabriel, 2 pages, 0.05 Mb

, pages, 0 Mb
Copyright © 2018 doc.uments.com