Social Indicators Research

, Volume 89, Issue 3, pp 565–571

What Do Happy People Do?


    • Department of SociologyUniversity of Maryland
  • Steven Martin
    • Department of SociologyUniversity of Maryland

DOI: 10.1007/s11205-008-9296-6

Cite this article as:
Robinson, J.P. & Martin, S. Soc Indic Res (2008) 89: 565. doi:10.1007/s11205-008-9296-6


Little attention in the quality-of-life literature has been paid to data on the daily activity patterns of happy and less happy people. Using ratings-scale information from time-diary studies, this article examines the hypothesis that people who describe themselves as happier engage in certain activities more than those who describe themselves as less happy. Based on 34 years of data collected by the General Social Survey (GSS) on social activities and media usage, it is found that people who are happy report being more active in most social activities, in religion and in newspaper reading. On the other hand, happier people report less time watching television, a relation that holds after control for education, marital status and other predictors of happiness. The need to replicate these findings using panel data is highlighted.


HappinessQuality of lifeTime useGeneral Social Survey (GSS)

1 Introduction

Google “happiness” and “activity” and you get quite a variety of entries and ideas to ponder. There is a lot of mention of Aristotle, and his confusing take on how to achieve it. There is mention of dogs, art, exercise and even a few empirical studies that suggest more active people are happier. Surveys showing political conservatives to be happier than liberals prompted syndicated columnist George Will to suggest his readers become politically conservative, thereby purging their souls of the curse of having a bleeding heart about society’s inequalities.

There are also a bevy of self-help books that suggest other ways to reach the exalted state of happiness. However, little of it is empirically based. Even Bradburn and Caplovitz’s (1965) classic tome on the topic, “Reports on Happiness,” concentrated more attention on the demographic correlates (age, marital status, etc.) of happiness rather than on its relation to activities (Phillips 1967).

This article examines some very comprehensive and long-run relations between happiness and activity from the 35+ year compilation of public opinion data from the General Social Survey (GSS). During that time, GSS has collected about 90 min worth of information from more than 45,000 people. Not all GSS respondents were asked questions about their activity, and it will be seen that the activities are not that numerous (10) and are mainly about things done in one’s free time (social life, TV, reading, religion). However, the GSS is an authoritative source with an accumulated large sample size over a long and significant period of American life.

The main GSS question asked about happiness has also been the source of much skepticism, giving respondents only the options of saying that they are “very,” “somewhat” or “not” happy; and only about 10% give the latter answer, leading to a limited measure of summarizing a complicated, and perhaps momentary and fleeting, reflection of the quality of one’s life. However, the question does provide some crucial hints about what the things we do have to do with more or less positive approaches to life.

2 More Enjoyable Daily Activities

As noted, there are only a few activities covered in the GSS, but there is an increasing body of data from “time-diary” studies that cover the complete spectrum of daily activity. Table 1 reviews some earlier data from diary studies done in years 1975 and 1985, both national probability surveys that first examined how people rated their day’s activities. These should give an initial starting perspective on what activities separate life’s ups from its downs.
Table 1

Enjoyment ratings from diary activities versus in general (1985 and 1975 national data, from Robinson and Godbey 1999, Appendix O)

1985 Diary (average = 7.0)

1975 General (average = 6.8)


9.3 Sex


9.2 Play sports


8.7 Playing/reading with children

8.9 Child care

8.5 Church, religion


8.5 Sleep

8.6 Play with children

8.2 Meals away

8.0 Socializing, talking

8.2 Socialize, visit others

8.0 Work

8.0 Socialize with family


8.0 Work breaks


7.9 Reading

7.5 Sleep

7.8 Meals at home

7.4 Eating

7.8 TV

7.4 Washing, dressing

7.4 Hobbies, crafts

7.3 Church, religion

7.2 Exercise

7.0 Reading

7.2 Baby care


7.2 Organizations


7.0 Work


7.0 Bathing


6.6 Cooking

6.8 Hobbies

6.6 Other shopping

6.5 Play sports

6.4 Child care

6.5 Cultural events

6.4 Help others

6.2 Cooking

6.3 Work commute


6.1 Dressing


5.8 Other housework

5.9 TV

5.5 Grocery shopping


5.5 Home repairs

5.1 Home repairs

5.2 Pay bills, financial etc.

5.0 Organizations

5.0 Yardwork


4.9 Clean house

4.6 Grocery shopping

4.9 Laundry

4.3 Other shopping

4.8 Health care, doctor

4.2 Clean house

4.7 Car repair



It first needs to be noted that the two studies employed different methodologies for rating daily activity. The first 1975 University of Michigan study asked respondents how much they enjoyed the specific activities on the right side of Table 1 (“Work,” “Sleep,” etc.) in general. The second 1985 University of Maryland study, shown on the left side of Table 1, used the same 0–10 enjoyment scale, but administered in “real time,” that is as it was experienced while they were engaged in it during their diary day. There are ratings for more activities in the 1985 study, because more of them were mentioned, or mentioned in more detail, in the context of the diary (like sex or work breaks). The diary ratings may also differ because they are being done by “doers” rather than the entire sample. Thus, a sports participant or organizational leader will probably rate that activity higher when they are doing it, compared to someone who doesn’t play sports or attend meetings, thus rating them lower in enjoyment on the general scale.

It can be seen in Table 1 that the two methods still correlate highly (.70), pointing to many of the same conclusions. For example, activities with children and socializing with others rated near the top in both scales, versus housework along with grocery shopping near the bottom. At the same time, even those activities rated near the bottom are not rated that negatively (about 5 on a 0–10 scale), so that they do not seem to be that much of a drag on daily life (It should be noted that the diary activity ratings are almost identical for men and women, with men rating household tasks lower than women in their general ratings).

What is highlighted in Table 1 are two important daily activities that go in opposite directions depending on the diary versus general time perspective. Work, for example is rated near the top (8.0) of activities in general. However, it falls toward the lower half (7.0) of activities in diary real time, suggesting that it has an aura of enjoyment in general (perhaps because one is glad just to have a job or social connections, offsetting the misery and ambiguity of unemployment) that conceals the everyday work disappointments and disasters at the workplace one experiences on a day-to-day basis.

On the other hand, TV is rated very low in general, but above average—and above work—in the diary report for the previous day. Put in other words, Table 2 suggests people’s underlying response might be, “TV programs in general are a waste of time, but the shows I saw last night were pretty good.” This distinction becomes important, because it will be seen below that TV is also a pivotal correlate of GSS happiness in general.
Table 2

Differences in occasions (per day, week, year) by happiness (GSS data 1972–2006—MCA-adjusted raw bivariate data in parentheses)



Very happy

Somewhat happy

Not happy

Difference (very-not)

Beta (eta)a

Socialize with (X/year):


Relatives (28,227)

80 (78)

7 (73)

69 (72)***

+11 times/year

.05 (.06)

Neighbors (28,701)

51 (46)

48 (49)

44 (49)**

+7 times/year

.03 (.03)

Friends (28,712)

51 (50)

47 (48)

46 (47)*

+5 times/year

.02 (.03)

Bar (28,667)

19 (16)

21 (21)

20 (22) NS

−1 time/year

.01 (.02)

Religious services (45,829)

25 (26)

20 (20)

18 (17)***

+7 times/year

.11 (.14)

Sex (5,618)

64 (64)

57 (57)

54 (50)***

+10 times/year

.06 (.06)

Newspaper days (15,572)

4.5 (4.6)

4.3 (4.3)

3.9 (3.8)***

+0.6 days/week

.05 (.05).

Internet hours/week (5,005)

6.3 (6.1)

6.5 (6.6)

6.7 (6.9) NS

−0.4 hours/week

−.01 (−.02)

TV hours/day (15,944)

2.7 (2.6)

2.9 (2.9)

3.5 (3.5)***

−0.8 hours/week

−10 (−.10)

Work hours/week (24,380)

41 (41)

41 (41)

40 (40) NS

+1 hour/week

.01 (.02)


59 (58)

61 (62)

49 (48)***

+10 points

.07 (.06)

aCorrelation after MCA adjustment (Beta) and before MCA (Eta), see text

*** Difference significant at <.001 level

** Difference significant at <.01 level

* Difference significant at <.05 level

For the other GSS activities, they also tend to be rated relatively highly, namely socializing, reading; and at the top of the diary ratings—sex. That gives us some expectation, then, that the more people engage in these activities, the happier they should be.

3 GSS Results

In the analysis below, for the sake of argument, comparability and simplicity, the tables are arranged as if activity is the dependent activity, with happiness as the predictor or independent variable. In other words, it looks like our implicit assumption is that happy people engage in these activities as a function of their happiness. It is recognized that the opposite may be just as or more plausible, that it is engaging in these activities makes people happier. Without panel data, however, it is hard to separate the chicken from the egg.

Thus, Table 2 is arranged to show differences in activity for people who are very versus somewhat versus not happy between 1972 and 2006. That makes it possible to say that happy people are 10% or 50% more (or less) active than less happy people, a handy common metric across activities. The basic numbers shown thus reflect level of activity (usually in estimated occasions per year, or hours per day or week). These numbers have been adjusted using the Multiple Classification Analysis (MCA) multiple regression program available as part of the ANOVA package in SPSS. MCA is here used to control for respondent age, education and marital status, each important predictors of both happiness and activity. Since many of these variables are correlated with each other, MCA adjusts for these correlations, revealing differences that are associated with this variable alone. The numbers in parentheses refer to the original, unadjusted bivariate differences, before MCA adjustment.

First, it can be seen that the expectation from Table 1 that socializing and happiness are related is confirmed. In Table 2, happier people do socialize more with relatives and friends (but less with neighbors), while they go to bars only slightly less. Similarly, very happy people go to church five times more a year than those who are somewhat happy and seven times more than not happy people. Moreover, happy people do engage in more sex, although not that much more frequently than the less happy, after adjustment for marital status and other demographics.

Turning to media activities, a similar positive pattern is found for newspaper reading, which is only slightly affected by demographic predictors. Almost no differences are found for Internet use. However, the pattern for daily TV use is particularly dramatic, with not happy people estimating over 30% more TV hours per day than very happy people, with somewhat happy people in between, with about 10% more hours than the very happy.

Finally, at the bottom of Table 2, it can be seen that work hour differences by happiness are minimal, all three groups working a little over 40 h a week. However, that is among those who are employed. Among those who are not happy, employment is 10–14 points lower than among the very or somewhat happy. As surmised in Table 1, being employed does relate to more positive feelings.

It should be noted that much the same pattern of results emerges from an analysis of the parallel GSS question on happiness with one’s marriage, again after MCA adjustment for demographic factors. As with personal happiness, respondents happier with their marriage visit more, particularly with relatives, and they go to bars less often than the not happy. They engage in 30% more sex, and they attend religious services more and read newspapers on more days. While those not happy with their marriages do watch more TV, it is only about 10% more than those with happy marriages.

4 Pace of Life

Finally, GSS has asked two perceptual questions about the pace of one’s life, as well as about specific activities. These questions ask how often respondents feel (1) rushed for time and (2) that they time on their hands they don’t what to do with. Tables 3 and 4 show that not happy people are different in feeling both more rushed for time and having too much time on their hands. Of the two, having extra time on one’s hands is the bigger burden of the two, with 51% of not happy people feeling that way, compared to only 19% of very happy people, with those somewhat happy people being about in the middle at 38%. Not happy people are also more likely to feel always rushed for time (35%) compared to 23% for those who are very happy. Of the two time dysfunctions, then, it seems better to feel rushed than to be adrift.
Table 3

GSS time perceptions by happiness



Very happy

Somewhat happy

Not happy

Difference (points)

Beta (eta)

(a) Rushed to do things


.05 (.07)

    Always rushed












    Almost never










(b) Time on hands


.15 (.19)

    Quite often






    Now and then






    Almost never










** Difference significant at <.01 level

Table 4

Happiness by time perceptions (% very happy)

Time on hands (n=)






Always rushed (891)



Sometimes rushed (1080)




39 (981)


Almost never rushed (410)




55 (156)


Nonetheless, these results remind us that too much activity can be dysfunctional, as a caveat that the general Table 2 finding that more activity is associated with a happier life style.

5 Summary and Conclusions

This research note has examined the relation between engaging in certain activities as a way of expressing or receiving happiness. Consistent with the idea that happy people engage in activities they rate as more enjoyable in time-diary studies, happier GSS respondents engaged in significantly more social activities, religious participation and newspaper reading. These differences held after important demographic predictors of activity were controlled.

The major exception was for TV viewing, which was the only GSS activity to correlate significantly lower with happiness. This reinforces conclusions from earlier studies of TV’s dysfunctional role in American society, both empirical (Kubey and Csikszentmihalyi 1990) and non-empirical (Mander 1978; Postman 1985). It stands in marked contrast to the high ratings TV scores as a daily activity in “real time,” not only in Table 1, but in the recent and more comprehensive studies done at Princeton University (Kahneman and Krueger 2006). If interpreted in the Princeton group’s framework of activity as experienced being the sine qua non of measurement, that would mean that TV represents a highly enjoyable activity that would improve the quality of people’s lives, given that more of Americans’ free time is being devoted to it.

Clearly, the data analyzed here point in the opposite direction. As noted at the outset, whether that means happiness leads to lower viewing, or that more viewing leads to unhappiness, cannot be determined from these data, and thus will require a panel design along with some careful observational study.

If we consider the alternative version of Table 2, with happiness as the outcome variable and activity as its predictor, one can see television as a predictor of general unhappiness. This could lead us to two possible interpretations:
  1. 1.

    Television viewing is a pleasurable enough activity with no lasting benefit, and it pushes aside time spent in other activities—ones that might be less immediately pleasurable, but that would provide long-term benefits in one’s condition. In other words, television does cause people to be less happy.

  2. 2.

    Television is a refuge for people who are already unhappy. TV is not judgmental nor difficult, so people with few social skills or resources for other activities can engage in it. Furthermore, chronic unhappiness can be socially and personally debilitating and can interfere with work and most social and personal activities, but even the unhappiest people can click a remote and be passively entertained by a TV. In other words, the causal order is reversed for people who watch television; unhappiness leads to television viewing. This argument makes it possible to discuss it in Table 2.


In tandem, these points have parallels with addiction; since addictive activities produce momentary pleasure but long-term misery and regret. People most vulnerable to addiction tend to be socially or personally disadvantaged, with TV becoming an opiate.

Given the time that Americans devote to it, the question deserves a clearer answer that should be answerable with panel data.

Copyright information

© Springer Science+Business Media B.V. 2008