Happiness, Affluence, and Altruism in the Postwar Period

Frank Levy

1.1 Happiness, Affluence, and Altruism

If an economist were asked to assess U.S. postwar economic performance, he would probably describe the period in three intervals:

The late 1940s and 1950s were fairly good. Real wages grew very quickly. The period had three recessions but only the last one (1958–60) was severe. Inflation was low except when WW II price controls were lifted and when the Korean War began. When inflation was a problem, a year of recession was sufficient to end it.

The 1960s were better. Real wages did not grow quite as fast in the 1950s but recessions were less of a problem. After 1963, the economy went on a sustained expansion. Unemployment fell more or less continuously until 1970. And inflation did not really emerge until the end of the decade.

The 1970s were awful. The 1960s expansion left an inflationary inertia so that the 1970–71 recession was not enough to bring inflation under control. Then came the 1972–73 food price explosion and the first OPEC price rise. They generated supply-shock inflation that was even more immune to recession. After 1973, productivity stopped growing and real wages stagnated. And then there was another OPEC price rise to finish out the decade. It was terrible.

The assessment appears non-controversial, yet it contains a large piece of thin ice. Terms like pretty good, better and awful imply something about not only the economy but about how people reacted to the economy. One could read the assessment as saying that people were happier in the 1960s than in the 1950s and least happy in the 1970s.

The most direct evidence of this proposition does not offer strong support. Periodically, the Gallup poll (officially the American Institute of Public Opinion — hereafter AIPO) asks questions with the following general form (e.g., AIPO 410): In general, how happy would you say that you are — very happy, fairly happy, or not very happy?

Responses to this question are contained in table 1.1. Data for the early postwar years through 1957 show a moderately increasing level of happiness — a possible reflection of rising real incomes. Then a six-year gap occurs during which time the question was not asked. When the data resumes in 1963, a trend is harder to discern. Interpretation is difficult because the precise wording of the question changes in 1963, 1973, and again in 1977. Moreover the 1977 response — the happiest in the series — comes from a poll that focused on religious habits and beliefs; responses for other years come from polls that focused on politics and the economy. When adjustments are made for these problems, the data suggest that the moderately rising level of happiness through the 1950s was followed by a roughly constant level of happiness thereafter.

Ten years ago, Richard A. Easterlin (1974) wrote an ingenious essay interpreting these poll responses in the context of James Duesenberry's relative income hypothesis (Duesenberry 1952). Easterlin began by noting that within any poll, higher-income individuals were more likely than lower-income individuals to report themselves as happy. He contrasted this association with his perception of a weaker association over time when real incomes were rising for everyone. He also examined data from Cantril's cross-national study (Cantril 1965) which showed a similar lack of association between a country's per capita income and the self-reported happiness of its population.

Together these data provided Easterlin with a basis for an application of the relative income hypothesis. In the application, an individual's happiness depends on the relationship between his income and his needs, but his needs are heavily conditioned by what he sees around him. If incomes were to rise uniformly, an individual's relative position (apart from lifecycle considerations) would remain unchanged, and so his individual happiness would not increase.

Easterlin's argument is appealing but it raises two problems. First, it does not explain the rising level of happiness in the early postwar years. Second, it leads to an overemphasis of private, versus public, consumption. As Easterlin writes:

Finally, with regard to growth economics, there is the view that the most developed economies — notably the United States, have entered an era of satiation. ... If the view suggested here has merit, economic growth does not raise a society to some ultimate state of plenty. Rather, the growth process itself engenders ever-growing wants that lead it ever onward. (1974, p. 121)

In the Easterlin-Duesenberry argument "ever-growing wants" refers to additional private consumption. It is private consumption, after all, that provides one's easiest comparisons with one's neighbors. But the focus on private consumption ignores important history.

The largest omission is the growth of the public sector. In 1947, all government nondefense outlays accounted for 14 percent of GNP. By 1980 these outlays had grown to 28 percent of GNP. In explaining this growth, the mid-1960s emerge as a particularly pivotal period during which the federal government instituted health insurance for the aged, the War on Poverty, aid to elementary and secondary schools (with emphasis on compensatory education) and other areas, which significantly redefined the role of the public sector.

To be sure, these Great Society years were distinguished not so much by growing expenditures as by new initiatives that would obligate future expenditures. Administration officials occasionally acknowledged the problems they were creating for future administrations. But they felt they had a rare opportunity, a narrow window, during which they had to finish what the New Deal had left undone (Moynihan 1967; Sundquist 1968).

Their opportunity came from a particularly sympathetic public. When government officials of the time proposed a new initiative, they emphasized its "public good" aspects: the way in which aiding the poor, the elderly, or the disadvantaged would make the United States a more humane place for everyone. Consider, for example, Lyndon Johnson's eloquent Howard University speech on equality for blacks:

... There is no single easy answer to all of these problems.

Jobs are part of the answer. They bring the income which permits a man to provide for his family.

Decent homes in decent surroundings, and a chance to learn — an equal chance to learn — are part of the answer.

Welfare and social programs better designed to hold families together is part of the answer.

Care of the sick is part of the answer.

An understanding heart by all Americans is also a large part of the answer.

To all these fronts — and a dozen more — I will dedicate the expanding efforts of the Johnson Administration.

... This is American justice. We have pursued it faithfully to the edge of our imperfections. And we have failed to find it for the American Negro.

It is the glorious opportunity of this generation to end the one huge wrong of the American Nation and, in so doing, to find America for ourselves, with the same immense thrill of discovery which gripped those who first began to realize that here, at last, was a home for freedom. (Quoted in Rainwater and Yancey 1967, pp. 131–32)

The rarity was not in Johnson's argument, but in the number of people who agreed with it. While the majority of the population did not demand such initiatives, they did form, in V. O. Key's phrase, a "permissive consensus" that allowed the government to implement its liberal agenda (Key 1961, p. 33). This willingness to experiment with public consumption (in hopes of increasing the general welfare) is not contained in Easterlin's reading of Dusenberry's theory. In this paper, we view the origins of the mid-1960s consensus from a somewhat different perspective.

A sensible explanation of the mid-1960s must account for both the origins of public consensus and its subsequent demise. By most estimates, the consensus for new initiatives peaked in 1964–66 and then began to erode. Government, acting in part on inertia, produced occasional new programs through the late 1960s and early 1970s, with the universalization of food stamps (1971) and the federal take-over of aid to the aged, blind, and disabled (1972) marking the end of the period. The remainder of the 1970s was increasingly dominated by antigovernment and antitax sentiment.

History affords few natural experiments; thus it is not surprising that the mid-1960s consensus has attracted a variety of explanations. The origin of consensus has been ascribed to the growth of the civil rights movement, and to the combination of John Kennedy's assassination and Barry Goldwater's candidacy (Sundquist 1968). The end of consensus has been ascribed to the devisiveness of the Vietnam War and to Watergate. But while all of these explanations sound plausible, none by itself is sufficient.

Consider, for example, the combined effects of the Kennedy assassination and the Goldwater candidacy. They were traumatic experiences that led to a Democratic president and Democratic congressional majorities. A rough parallel existed in 1976 when, in the aftermath of Watergate, the country again elected a Democratic president with Democratic congressional majorities. Despite the parallels, Jimmy Carter had only a fraction of the legislative success enjoyed by Lyndon Johnson. Part of Carter's lack of success was due to his political style, but part was due to the temper of the times: If people accepted "public good" arguments in the mid-1960s, they clearly rejected those arguments in the 1970s. They had seen too many government programs in the intervening years to retain their enthusiasm.

This dependence of current attitudes on past experience leads us to look for a more generic explanation of the mid-1960s, an explanation that (like Easterlin's explanation of happiness) does not rest too heavily on particular presidents, wars, or scandals.

One such explanation involves a slight elaboration of a public-private cycle recently advanced by Albert Hirschman. In Shifting Involvements (1979), Hirschman argues that industrialized societies move between periods emphasizing private material gain and periods emphasizing public (i.e., collective) action. The cycle's motivation is the disappointment that arises when either kind of "consumption" fails to provide the satisfaction it promised ex ante.

In his writing Hirschman discusses government programs under the heading of private material gain (pp. 39–46). But his examples make clear that he is describing those government programs that are most like private goods — the public education a father and mother "purchase" for their children. Conversely, the redistributive programs of the mid-1960s are best thought of as government extensions of Hirschman's collective action.

With this modification, the Hirschman cycle begins with a period of rising consumption and rising happiness. After a time the attraction of increasing affluence begins to wear thin. This leads to the cycle's second phase in which popular opinion becomes receptive to both public causes and those government programs that advance the "public good." (Again, it is the public or collective nature of such programs that is important. A war against poverty, properly presented, can fit the category. A war against potholes in one's own neighborhood cannot.)

If the government takes up this mandate, the public will find that the country is not improved as easily as they had hoped (or as the government had promised). Their disappointment in public goods — the Hirschman cycle's third phase — leads to a returning emphasis on private affluence.

In this chapter we ask whether a Hirschman cycle lay behind the mid-1960s consensus. Given the limits of public opinion data — gaps in time series, questions with changed wording — the case for the cycle is reasonably strong. The result increases our understanding of the postwar American experience in two ways. Not only does it give us a better explanation of the mid-1960s consensus, but it helps to explain why the federal government could look so good in the mid-1960s and so bad thereafter. And, as a final point, we will find that the Hirschman cycle and the Easterlin-Dusenberry hypothesis are not so different as they first appear.

We will develop the argument in the three sections that follow. In section 1.2 we briefly review the macroeconomic history of the postwar period and public reaction to economic conditions. We conclude that people's economic optimism rose through the 1950s and that by the late 1950s they appear to have put the uncertainty of the Great Depression, World War II, and the Korean War behind them. This trend is consistent with the trend of increasing happiness during the period (table 1.1) and suggests that if a Hirschman cycle existed, support for public goods should have grown at about this time.

The economic data also show that people became uncertain about the economy at the end of the 1960s; the data became particularly gloomy after the food and oil price inflation of 1973–74. If a Hirschman cycle existed, disappointment with public goods could have begun well before this time, but adverse economic conditions would have reinforced the turn back to private consumption.

In section 1.3 we review opinion data for the existence of a Hirschman cycle. In the late 1950s, the most obvious examples of broad public goods were redistributive programs. We find that during this period, sentiment in favor of such programs was high and increasing. Initially this sentiment did not translate into new programs as Presidents Eisenhower and Kennedy were reluctant to propose substantial new spending. Their reluctance, if anything, made people more amenable to Lyndon Johnson's public good arguments because there were few recent examples of public good programs that had failed. It was under Johnson that substantial new initiatives began, and our data show that not long afterwards sentiment in favor of redistributive programs began to dissipate.

In section 1.4 we summarize our findings and apply them to understanding the role of government in the 1960s and 1970s. We conclude by briefly reconciling the Hirschman cycle with the Easterlin-Dusenberry argument.

1.2 The Postwar Economic Landscape

The experience of the Great Depression and rationed wartime consumption dictated that people would enter the postwar period with a hunger for private consumption. The question is, when, if ever, did this hunger abate: Did private consumption grow enough and become sufficiently secure to cause public opinion to shift toward collective welfare as Hirschman's cycle predicts?

We begin to answer this question by briefly reviewing the economics of the postwar period, with particular emphasis on the path of incomes. In the early years of the period, the dominant single factor was rising productivity. Between 1947 and 1959 the growth of output per manhour in private business averaged 3.3 percent per year compared to 2.5 percent per year for both the 1910–29 and 1930–39 periods. The high growth rate was important because it translated into rapidly rising real wages (table 1.2).