BOWLING ALONE

by Robert D. Putnam

  31. Sourcebook of Criminal Justice Statistics—1995, ed. Kathleen Maguire and Ann L. Pastore (Albany, N.Y.: Hindelang Criminal Justice Research Center, 1996), 365. See also U.S. Public Health Service, The Surgeon General’s Call to Action to Prevent Suicide (Washington, D.C.: 1999), and the research cited there. On the low suicide rates among those I term “the long civic generation,” see Max A. Woodbury, Kenneth G. Manton, and Dan Blazer, “Trends in U.S. Suicide Mortality Rates 1968 to 1982: Race and Sex Differences in Age, Period and Cohort Components,” International Journal of Epidemiology 17 (1988): 356–362, esp. 360. For broadly comparable patterns in other nations, see C. Pritchard, “New Patterns of Suicide by Age and Gender in the United Kingdom and the Western World 1974–1992: An Indicator of Social Change?”Social Psychiatry and Psychiatric Epidemiology 31 (1996): 227–234.

  32. Michael Rutter and David J. Smith, “Towards Causal Explanations of Time Trends in Psychosocial Disorders of Young People,” in Psychosocial Disorders in Young People, Rutter and Smith, eds., 807.

  33. On our index of malaise, see note 50, chapter 13. Some symptoms of malaise show life cycle effects—sleeplessness slightly increases with age, while headaches decrease with age—but life cycle differences are screened out of figure 74. “High” malaise means in the top third of all respondents over these twenty-five years, but any reasonable cut point would produce the same results. Financial worries have grown in the younger cohort over the last quarter century, and financial worries in turn produce headaches, indigestion, and sleepless nights. When our index of financial worries is added to a multiple regression prediction for malaise (including sex, education, age, reliance on TV for entertainment, and an interactive term for age and year), the unstandardized regression coefficient on the interactive term—a statistical measure of the growing generation gap—is cut by roughly 60 percent but remains highly significant.

  34. Ed Diener, “Subjective Well-Being,” Psychological Bulletin 95 (1984): 542–575, esp. 554. W. A. Stock, M. A. Okun, M. J. Haring, and R. W. Witter, “Age and Subjective Well-being: A Meta-analysis,” in R. J. Light (ed.), Evaluation Studies: Annual Review, vol. 8 (Beverly Hills, Calif.: Sage, 1983), 279–302; D. D. Witt, G. D. Lowe, C. W. Peek, and E. W. Curry, “The Changing Relationship between Age and Happiness: Emerging Trend or Methodological Artifact?” Social Forces 58 (1979): 1302–1307; and author’s analysis of DDB Needham Life Style archive. For our index of life contentment, see chapter 20.

  35. Schneider and Stevenson,Ambitious Generation, 189–211, quotation at 192; Seligman, “Boomer Blues,” 52, 55.

  36. L. I. Pearlin, M. A. Lieberman, E. G. Menaghan, and J. T. Mullan, “The Stress Process,” Journal of Health and Social Behavior 22 (1981): 337–356; P. Cohen, E. L. Struening, G. L. Muhlin, L. E. Genevie, S. R. Kaplan, and H. B. Peck, “Community Stressors, Mediating Conditions and Wellbeing in Urban Neighborhoods,” Journal of Community Psychology 10 (1982): 377–391; A. Billings and R. Moos, “Social Support and Functioning among Community and Clinical Groups: A Panel Model,” Journal of Behavioral Medicine 5 (1982): 295–311; Nan Lin and W. M. Ensel, “Depression-Mobility and Its Social Etiology: The Role of Life Events and Social Support,” Journal of Health and Social Behavior 25 (1984): 176–188; G. A. Kaplan, R. E. Roberts, T. C. Camacho, and J. C. Coyne, “Psychosocial Predictors of Depression,” American Journal of Epidemiology 125 (1987): 206–220.

  37. This generalization summarizes the author’s extensive multivariate analysis of dozens of measures of civic engagement and social capital in the Roper Social and Political Trends survey archive, the DDB Needham Life Style survey archive, the General Social Survey, the National Election Studies, the Americans’ Use of Time archive, the Monitoring the Future archive, and others. (Generational analysis of the Roper Social and Political Trends surveys is seriously complicated by the fact that “age”—and thus year of birth—is crudely measured in these surveys, so regression analysis of these data provide less clear-cut support for the generational interpretation. On the other hand, see table 3 for evidence of the role in generation in the Roper data.) The central question in these analyses was this: By what fraction is the trend over time (the unstandardized regression coefficient for year of survey, for example) reduced when generation is controlled (by entering year of birth in the regression, for example)? As discussed earlier, for some indicators of civic engagement—voting, church attendance, newspaper readership, interest in public affairs, and social trust—virtually all net change over the last third of the twentieth century is attributable to generational change. This pattern can be seen, for example, in figures 39 and 53 and in the fact that if both year of birth and year of survey are included in the same regression, year of survey becomes virtually insignificant as a predictor of these measures. For other indicators of social capital, like club meetings and family dining, somewhat less than half of the trend is eliminated when generation is controlled. For some measures of schmoozing, such as playing cards and entertaining at home, controls for generation have little or no effect on the trends.

  38. William Graham Sumner, Folkways: A Study of the Sociological Importance of Usages, Manners, Customs, Mores, and Morals (Boston: Ginn, 1911), 12–13; Lewis A. Coser, The Functions of Social Conflict (Glencoe, Ill.: Free Press, 1956); Arthur A. Stein, “Conflict and Cohesion,” Journal of Conflict Resolution 20 (1976): 142–172; Theda Skocpol, Ziad Munson, Marshall Ganz, and Andrew Karch, “War and the Development of American Civil Society,” paper prepared for annual meeting of the American Sociological Association (Chicago, August 1999); Susan J. Ellis, and Katherine H. Noyes, By the People: A History of Americans as Volunteers, rev. ed. (San Francisco: Jossey-Bass, 1990), quotation at 13.

  39. Charles, Service Clubs, 15–16, 31.

  40. Thanks to Wendy Rahn and Theda Skocpol for illuminating discussions of the effects of war, especially World War II, on social capital and civic engagement. See Theda Skocpol, with the assistance of Marshall Ganz, Ziad Munson, Bayliss Camp, Michele Swers, and Jennifer Oser, “How Americans Became Civic,” in Civic Engagement in American Democracy, eds. Skocpol and Fiorina, 27–80, and Tom Brokaw, The Greatest Generation (New York: Random House, 1998).

  41. John Morton Blum, V Was for Victory: Politics and American Culture during World War II (New York: Harcourt, Brace, Jovanovich, 1976), 339; author’s analysis of the General Social Survey (1974–94) and the DDB Needham Life Style archive (1983–88). Veterans are not more engaged civically than other men of their generation. The enduring effects of World War II on the civic habits of those who lived through it were not limited to the battlefield. Or perhaps the brutalizing effects of combat counterbalanced its communitarian effects.

  42. Richard R. Lingeman, Don’t You Know There’s a War On? The American Home Front, 1941–1945 (New York: G. P. Putnam’s Sons, 1970), 71; Bill Gold, quoted in Roy Hoopes, Americans Remember the Home Front: An Oral Narrative (New York: Hawthorne, 1977), xii.

  43. Richard Polenberg, War and Society: The United States, 1941–1945 (New York: J. B. Lippincott, 1972), 17.

  44. Polenberg, War and Society, 29–30.

  45. The Crosby jingle appears in a taped collection of wartime memorabilia, The Home Front, 1938–1945 (Petaluma, Calif.: The Mind’s Eye, 1985).

  46. Lingeman, Don’t You Know, 237, estimates 335,000 tons; Polenberg, War and Society, 16, suggests 450,000 tons. The president’s appeal is quoted at Polenberg, War and Society, 16.

  47. Lingeman, Don’t You Know, 52, 59, 62, 250; Red Cross national membership records.

  48. Lingeman, Don’t You Know, 251.

  49. Julie Siebel, “Silent Partners/Active Leaders: The Association of Junior Leagues, The Office of Civilian Defense, and Community Welfare in World War II” (Ph.D. diss., University of Southern California, 1999).

  50. Polenberg, War and Society, 132, citing W. Lloyd Warner, “The American Town,” in American Society in Wartime, William Fielding Ogburn, ed. (Chicago: University of Chicago Press, 1943), 45–46.

  51. Jeffr
ey G. Williamson and Peter H. Lindert, American Inequality: A Macroeconomic History (New York: Academic Press, 1980), esp. 53–54 and 82–92, data at 54 and 315. See also Polenberg, War and Society, 94. World War I had similarly sharply reduced economic inequality, but the equalizing effect of that war vanished within a year or two, whereas the more egalitarian distribution of wealth and income after World War II persisted and even improved until the early 1970s.

  52. Polenberg, War and Society, 137. In retrospect one might be surprised that as much as 80 percent of Americans believed that using the black market was never justified.

  53. Polenberg, War and Society, 140–145, quotation at 143. Brian M. Downing, The Paths of Glory: War and Social Change in Twentieth-Century America (forthcoming, 2000), argues that the war’s disruptive effects of community outweighed its positive effects.

  54. Personal communication, Robert Rosenheck, M.D. (New Haven, Conn., Veterans’ Administration).

  55. Blum, V Was for Victory, 340.

  56. Thanks to Professor Rahn for the data for figure 75, drawn from a July 1998Wall Street Journal /NBC News Poll.

  57. Author’s analysis of Roper Social and Political Trends archive through 1991, augmented for 1994 and 1996 from the relevant Roper Reports (New York: Roper Starch Worldwide, various years). “Material luxuries” in figure 76 refers to respondents who chose at least two of the following six items as part of the definition of “the good life”: a job that pays more than average, a swimming pool, a vacation home, really nice clothes, a second color TV set, a second car. Controlling for income, education, marital status, sex, and city size, both year of survey and year of birth are highly significant predictors of materialism, but year of birth (representing generational differences) is by far the strongest predictor.

  58. The alternatives offered were these: my family; my old friends; my new friends; the people in my neighborhood; my church/synagogue; the people I work with; my local community; reading local newspapers; the organizations or groups I belong to; parents of my children’s friends; reading special interest magazines; the people I meet on-line on the computer. More than one alternative could be chosen. To simplify figure 77, I have consolidated “old” and “new” friends and eliminated “parents of my children’s friends” and “special interest magazines.” Overall, 9 percent mentioned magazines and 28 percent of parents with children at home mentioned other parents; neither alternative differed significantly by generation. “Co-workers” is calculated on the basis only of respondents who work at least part-time outside the home. The cohort breakdown was determined by Yankelovich Partners and excludes respondents born after 1978; to simplify figure 77, I have omitted the baby boomers (born 1946–64); almost without exception they fall midway between the other two cohorts. There are no significant differences among the surveys in 1997, 1998, and 1999; figure 77 presents the average for these three years. Thanks to Yankelovich Partners for making these data available.

  59. William James, “The Moral Equivalent of War” (New York: American Association for International Conciliation, 1910).

  CHAPTER 15: WHAT KILLED CIVIC ENGAGEMENT? SUMMING UP

  1. Theodore Caplow, Howard M. Bahr, John Modell, and Bruce Chadwick, Recent Social Trends in the United States: 1960–1990 (Montreal: McGill-Queen’s University Press, 1991), 47, 106, 11; U.S. Bureau of the Census, Current Population Reports, Series P20-509, “Household and Family Characteristics: March 1997,” and earlier reports. Author’s analysis of General Social Survey.

  2. All generalizations in the preceding paragraphs are based on the author’s analysis of the Roper Social and Political Trends, DDB Needham Life Style, Americans’ Use of Time archive, General Social Survey, and National Election Studies archives, controlling for all standard demographic characteristics. This conclusion differs from my speculation in “Tuning In, Tuning Out” and is based on a much broader range of evidence on the links between family structure and social connectedness.

  3. Verba, Schlozman, and Brady, Voice and Equality, 241–247.

  4. Author’s analysis of General Social Survey, Roper Social and Political Trends, and DDB Needham Life Style archives. White support for segregation is measured by this question in the GSS: “If you and your friends belonged to a social club that would not let blacks join, would you try to change the rules so that blacks could join?” Similar results obtain if white racism is measured by support for residential segregation or antimiscegenation laws.

  5. Fukuyama, Trust, 313–314. On the debate about whether government programs “crowd out” philanthropy and volunteering and erode social capital, see Paul L. Menchik and Burton A. Weisbrod, “Volunteer Labor Supply,” Journal of Public Economics 32 (1987): 159–183; Susan Chambre, “Kindling Points of Light: Volunteering as Public Policy,”Nonprofit and Voluntary Studies Quarterly 18 (1989): 249–268; Richard Steinberg, “The Theory of Crowding Out: Donations, Local Government Spending, and the ‘New Federalism,’” in Philanthropic Giving, Richard Magat, ed. (New York: Oxford University Press, 1989), 143–156; Marvin Olasky, The Tragedy of American Compassion (Washington, D.C.: Regnery Gateway, 1992); Peter Dobkin Hall, Inventing the Nonprofit Sector (Baltimore, Md.: Johns Hopkins University Press, 1992), 1–83; Robert Moffitt, “Incentive Effects of the U.S. Welfare System: A Review,” Journal of Economic Literature 30 (1992): 1–61; Deborah Stone, “The Durability of Social Capital,”Journal of Health Politics, Policy, and Law 20 (1995): 689–694; and J. David Greenstone and Paul E. Peterson, Race and Authority in Urban Politics: Community Participation and the War on Poverty (New York: Russell Sage Foundation, 1973).

  6. Statewide differences in levels of social capital, as discussed in section IV, are substantial, closely intercorrelated, and reasonably stable, at least from the 1970s to the 1990s.

  7. Putnam, “Tuning In, Tuning Out,” 671.

  8. Daniel Bell, The Cultural Contradictions of Capitalism, 20th anniv. ed. (New York: Basic Books, 1996); Robert E. Lane, The Market Experience (New York: Cambridge University Press, 1991).

  9. Charles H. Heying, “Civic Elites and Corporate Delocalization: An Alternative Explanation for Declining Civic Engagement,” American Behavioral Scientist 40 (1997): 657–668.

  10. Another possible explanation for civic disengagement is the rising crime rate during the 1970s and 1980s. However, if we control for other influences on connectedness (education, race, income, generation, gender, marital, parental, and job status, financial worries, city size, homeownership, residential mobility, commuting time, and TV dependence), neither the objective crime rate in the surrounding county nor the subjective fear of crime is correlated with such measures of civic engagement as club meeting attendance, home entertaining, visiting friends, interest in politics, or engaging in community projects. I find no evidence that the civic disengagement described in section II is a result of increased crime.

  11. These rough estimates of the relative importance of various causal factors are derived from multiple regression analyses across all the major data sets in this study and all the major indicators of social and political participation. In effect, I asked, “How much would civic participation or social capital have declined if the relevant causal factor— the fraction of women in the workforce, economic anxiety, suburbanization, TV viewing, and so forth—had not changed over the last third of the twentieth century?” Necessarily, this approach abstracts from minor differences across various measures and assumes away any synergistic effects. However, as a general summary it does no violence to the underlying evidence.

  CHAPTER 16: INTRODUCTION

  1. Kenneth J. Arrow, “Gifts and Exchanges,” Philosophy and Public Affairs 1 (summer 1972): 357.

  2. The measure of group membership is from the General Social Survey as described in chapter 3 and is available for forty states. The measures of public meetings and local organizational leadership are from the Roper archive as described in chapter 2 and are available for forty-three states. The measures of club meetings, volunteering, and community p
rojects are from the DDB Needham archive as described in chapters 3 and 7 and are available for forty-eight states.

  3. The specific questions from the DDB Needham archive: “I spend a lot of time visiting friends” (agree-disagree) and “How often in the last year did you entertain at home?” They are available for forty-eight states.

  4. The specific questions are from the DDB Needham archive (“Most people are honest”), available for forty-eight states, and the General Social Survey (“Most people can be trusted” vs. “You can’t be too careful”), available for forty-one states. Though entirely distinct methodologically these two statewide measures of social trust are quite convergent (r = .79 for all available states; r = .85 for the thirty-eight states for which at least one hundred respondents are available in each survey).

  5. Our measure of turnout is simply the average percentage of the voting-age population who voted in the presidential elections of 1988 and 1992, as reported in the U.S. Statistical Abstract, 1994: 289. These data are available for all fifty states.

  6. Our measure of the incidence of nonprofit (501[c]3) organizations is simply the number of such organizations in each state in 1989 (as reported in the Non-Profit Almanac for 1992–93), divided by the state’s population in 1990. (I thank Professor Tom W. Rice for pointing me to these data.) This measure is stable over time; the 1989 measure that we use is very strongly correlated (r = .89) with the same measure in 1992. Our measure of the incidence of civic associations is the mean number of “civic and social associations” (SIC 8640) reported annually from 1977 to 1992 by the Commerce Department, divided by the state’s population in each year. Both sets of data are available for all fifty states.