CYCLES IN THE MASS PSYCHOLOGY OF AMERICAN ELECTORAL BEHAVIOR by Ted Goertzel, Rutgers University, Camden NJ 08102 internet: goertzel@camden.rutgers.edu voice: 609 429-4921 fax: 609 225-6495 Note: This paper is forthcoming in the journal Psicologia Politica (Valencia, Spain). Please mention this if you should cite the paper. There may be a few errors in this draft which were changed for the final Spanish draft, but I can't remember what they were to change them here. The election of Bill Clinton, reversing a 12 year period of dominance by conservative Republicans, has revived interest in cyclical theories of American political behavior. This paper reviews a number of theoretical traditions which try to explain cyclical phenomena in politics, culture and mass psychology. It looks for points where these theories differ and where they can be fruitfully combined, and considers the data which has been ad- vanced to support them. Several varieties of theories are examined: political, socio-cultural, ideological, economic, generational and psychological. However, the theories cannot be sorted neatly into these categories since many authors wisely incorporate several dimensions in their work. We will begin with some of the theories which stress elec- toral behavior. THE SCHLESINGERS. Arthur Schlesinger, Sr. (1949: 77-92) and Arthur Schlesinger, Jr. (1986: 23-50; 1992) have a family tradition of insightful commentary on historical cycles in American party politics. In their view the American populace is split between liberals and conservatives. Liberals seek to advance public purposes while conservatives seek freedom to pursue private interests. Shifts in the preponderance of opinion from one side to the other cause periodic changes in the national mood. They observe that these shifts have occurred at regular intervals throughout American history, as shown in Table One. TABLE ONE Liberal Conservative 1765-1787 1787-1801 1801-1816 1816-1829 1829-1841 1841-1861 1861-1869 1869-1901 1901-1919 1919-1931 1931-1947 1947-1961 1961-1978 1978-1993 1993-[2010] [2010-2026] Schlesinger, Sr., observed that the average length of the eleven periods up to 1947 was 16.55 years. The three periods since then have been of roughly the same magnitude: 14, 17 and 15 years. Voting in national elections is the Schlesingers' main source of data, which means that changes in the national mood may go undetected until they are revealed in an election year. A major national crisis, also, may disrupt the usual rhythm. The most impressive feature of the Schlesingers' work is their remarkable record of successful prediction. In 1924, Schlesinger, Sr., predicted that Coolidge-style conservatism would last until 1932. In 1939, he predicted that the liberal mood would end in about 1947. In 1949, he predicted a shift to liberalism in 1962 and a conservative shift in 1978. These predictions seem generally consistent with the trends which emerged, although not closely enough to enable them to predict specific elections. Schlesinger, Jr. published "How McGovern Will Win" in 1972. Bill Clinton's election in 1993 is an impressive validation of their model, going against the conventional wisdom in the late 1980s which thought that the period of conservative Republican dominance would be much longer. The Schlesingers' are not wedded to any particular theo- retical explanation of the trends which they observe. At various times, they have considered sun spot cycles, biorythms, and generational succession as possible causal factors. Most convincingly, however, they turn to the dynamics of political organization itself. It seems to take about fifteen years for a successful political party or movement to define its agenda, mobilize its resources, implement its policies as best it can, and obtain the inevitably less than hoped for results. Then the opposing tendency has its turn. KLINGBERG. Also in the political science tradition is Frank Klingberg (1952) who wrote about the alternation of moods in American foreign policy, without relating them to underlying psychological factors. However, the terms which he used to categorize the alternating moods, introversion and extroversion, are more psychological than political. During extroverted periods, the nation has been willing to exert economic, diplomatic and/or military power on other nations. during introverted periods, it has been unwilling to do so, perhaps because of a preoccupation with domestic concerns. Once a dominant mood is firmly established it persists for two decades or more. Since 1976 there have been four introvert phases averaging about 21 years in length, and three introvert phases averaging approximately 27 years in length. The most recent extroverted phase began in 1940 or 1941. Klingberg's article was published in the middle of this phase which seems to have ended approximately on time with the end of the Vietnam war in 1973. The period from 1973 to 1991 could be characterized as an introverted phase in American foreign policy, generally attributed to a fear of repeating the Vietnam debacle. President George Bush proclaimed the end of this "Vietnam Syndrome" with the intervention which drove Iraqi troops from Kuwait in 1991. Klingberg noted a similarity between his observations and those of Arthur Schlesinger, Sr., but his cycles are of longer duration and the turning points do not often coincide. This difference may have something to do with his focus on foreign policy, while the Schlesingers' focus is on domestic social and economic policy. His data base is also different. While the Schlesingers relied on qualitative descriptions of electoral cam- paign themes, Klingberg collected a good deal of quantitative data. He content analyzed Presidential "state of the union" ad- dresses, coding the percentage of the address which mentioned foreign relations or expounded the importance of positive action in foreign affairs. He studied party platforms and trends in naval expenditures. His extensive data is clearly reproduced in bar and line graphs which reveal the difficulty inherent in this kind of work. The patterns are irregular, with the long term trends obscured by short term fluctuations. This is inherent in the data, and Klingberg does not resort to any of the smoothing techniques which sometimes make these trends seem deceptively clear. A good deal of clinical judgment is required to fit this data into the clear patterns presented in his article, and there is room for disagreement about where one periods ends and another begins. TABLE TWO Introverted Extroverted 1776-1798 1798-1824 1824-1844 1844-1871 1871-1891 1891-1919 1919-1940 1940-1973 1973-1991 1991- Like the Schlesingers, Klingberg is not committed to any one theoretical explanation for the cycles which he has observed. He reviews a comprehensive list of possible causes: periodic exhaustion with any dominant policy and desire for something new, generational change, political or economic factors within the United States, changes in the policies of other nations, He expresses the view that the introversion-extroversion cycle may be more fundamental than the liberal-conservative cycle. He speculates that Toynbee's "challenge and response" model may be applicable, and suggests that American history can be interpreted as a series of responses to fundamental challenges: e.g., the revolutionary cycle, the anti-slavery cycle and the populist cycle. In 1952, he saw America as rising to the challenge of world leadership in defense of the free world. SOROKIN. Sociology's main contributor to cycle theory is Pitirim Sorokin whose four volume Social and Cultural Dynamics (1937-1942; abridged edition 1957) is perhaps the most ambitious study of cycles in world history. Sorokin had an elaborate theory with which he attempted to test with empirical data. He argued that history moved slowly between two systems of ideas: sensate and intuitive. The sensate system is empiricist, nominalist, determinist and materialist. It believes that the physical, material world is the fundamental reality. The ideational is idealist, mystical, and realist in the sense of believing that ideas have an existence independent of the material world. Sorokin argued that cyclical changes in a system are due largely to an "unfolding of its inherent possibilities" (Sorokin, 1941: 5). Sorokin applied his scheme to world history over very long periods of time, not to the short term cycles which are the focus of this paper. For example, he regarded the triumph of Christianity over the Roman empire as an ideational reaction against the excessively sensate nature of the Roman empire. His data was primarily intellectual works by leading philosophers and writers of the various eras. A careful reanalysis of his data by Dean Keith Simonton (1976, 1984) did not confirm Sorokin's conclusions. Simonton found that sensate and ideational philosophies do not alternate in time, but actually fluctuate together. Historical eras with intense philosophical and intellectual activity (both sensate and intuitive) are followed by periods of low productivity of either type. Of course, this analysis is based entirely on published writers and may not correspond to trends in mass consciousness. So far as I know, no one has attempted to apply Sorokin's theory to short term cycles in American culture or politics. STRAUSS AND HOWE. William Strauss and Neil Howe's Generations (1991) is a tour de force of generational analysis, covering the full sweep of the American past and future from 1584 to 2069. Their book incorporates and supersedes the work of several other generational analysts of American political history (Elazar, 1976; Keller, 1979; Huntington, 1974) whose historical observations are similar but less systematically developed. By developing the generational argument as fully as they do, they clarify its limitations as well as its strengths. The concept of generation works best within families, the biology of the human life cycle dictates that distinct generations succeed each other in 20 to 30 year intervals. In a society as a whole, however, births are continuous, and the division into generations is controversial. There do seem to be changes in the cultural and political zeitgeist from decade to decade (Mannheim, 1952; Goertzel, 1972), but historians do not always agree in characterizing these changes or in drawing lines between epochs. Strauss and Howe reconcile the life cycle and historical approach by defining a generation as "a special cohort-group whose length approximately matches that of a basic phase of life, or about twenty-two years over the last three centuries" (p. 34). They claim that a distinctive "social moment" has occurred every twenty-two years or so, marking the transition from one generation's period of dominance to that of another. They further argue that American history is marked by four generation cycles, which last approximately ninety years. This long cycle theory is what enables them to make predictions as far into the future as 2069. They argue that four generational types have recurred in a fixed order (with one exception) throughout American history. The four generation theory is not a new idea, but can be traced back to Exodus and The Illiad as well as to the works of Huntington (1981), Marias (1967, 1968), Littre (1860) and Ferrari (1872). Briefly summarized, the four cycles are as shown in the table. Strauss and Howe follow each of these generations through the four stages of the life cycle (youth [0-21], rising adulthood [22-43], midlife [44-65] and elder [66-87], and explore how the interpersonal dynamics within families tend to differ depending on which historical generation each cohort is part of. However, this discussion is too complex for this short paper, so I will concentrate on the midlife period which when each generation exercises its greatest political impact. TABLE THREE IDEALIST An inner-driven, moralistic generation which comes of age during a period of spiritual awakening and develops a new creedal passion. REACTIVE An alienated, cynical generation which challenges the ideals of their parents and develops into pragmatic, risk-taking adults. CIVIC An outer-driven, morally complacent generation which institutionalizes many of the ideals of the previous generations. ADAPTIVE A hypocritical generation which coasts along on the accomplishments of the civics, laying the groundwork for a new idealist era. The concept of "social moment" plays a key role in Strauss and Howe's theory. They postulate that each of their long cycles has two such moments, the first of which is a "spiritual awakening," the second a "secular crisis." These social moments are as follows: TABLE FOUR SPIRITUAL AWAKENING SECULAR CRISIS Reformation Awakening (1517- 1539) Defeat of Spanish Armada (1580-1588) Puritan Awakening (1621-1640) Glorious Revolution (1675- 1692) Great Awakening (1734-1743) American Revolution (1773- 1789) Transcendental Awakening (1822-1837) Civil War (1857-1865) Missionary Awakening (1886- 1903) Depression & World War II (1932-1945) Boom Awakening (1967-1980) [expected in the 2020's] This is certainly a grand historical scheme, very much in the tradition of Sorokin and other students of long term cultural trends. One can, of course, question whether these really were the defining events of American history, or whether they have been selected because they fit the scheme. Was the cultural revolution of the 1967-1980 period really the historical equivalent of the Great Awakening? Do the great depression and World War II really constitute one event? Why is World War I left out? One can also question whether the these events really determined generational succession in the zeitgeist. Strauss and Howe do rather well in fitting cultural history into their scheme, except for the post-Civil War period when they argue that no civic generation appeared because the Civil War itself happened too soon. While this scheme covers the full sweep of American history, the cycles are so long that only four and a half of them have occurred. This does not provide enough cases for any kind of statistical certainty, an inherent problem in long cycle theories. What can this kind of sweeping theory tell us about contemporary politics and culture? To answer this, we can look at the generations which are currently active politically, of which there are five. These generations are described in the following table. These generations do make a certain amount of cultural sense; the G.I. generation and the Baby Boomers in particular are well known as cultural phenomena. The Silent Generation does seem somewhat weaker politically, as shown by the failure to elect any Presidents. The election of Bill Clinton is widely interpreted as a coming to power of the Baby Boomers. Interestingly enough, however, Strauss and Howe did not predict the boomers coming to power in the 1992 elections. Indeed, they differed with Arthur Schlesinger, Jr. on this point, remarking that: In the spring of 1988, Arthur Schlesinger, Jr., published an article entitled "Wake Up Liberals, Your Time Has come," predicting a resurgence of sixties-style liberalism over the next few years. To date, Schlesinger's prediction seems wide of the mark. One problem, in our view, is that a sixteen-year pendulum is too exact--and (unlike Klingberg's twenty-three-year pendulum) speeds the clock too fast. (Strauss and Howe, 1992: 103). TABLE FIVE GENERATION (type) YEARS BORN (duration) PRESIDENTS & CANDIDATES G.I. (Civic) 1901-1924 (24 years) Johnson, Reagan, Nixon, Ford, Kennedy, Carter, Bush SILENT (Adaptive) 1925-1942 (18 years) Mondale, Dukakis, Kemp, Hart, Jackson BOOM (Idealist) 1943-1960 (18 years) Bill Bradley, William Bennett, Newt Gingrich, Dan Quayle, Albert Gore, [Bill Clinton] THIRTEENTH (Reactive) 1961-1981 (21years) None Strauss and Howe (1992: 105) concede that "the acid test of any theory is its ability to forecast," and Schlesinger seems to have won this one. However, they claim to be able to predict long term trends only, not specific events. More generally, however, Strauss and Howe are addressing a different kind of phenomena than the Schlesingers. The Schlesingers are interested in the alternation of power between liberals and conservatives. They would consider John F. Kennedy and Ronald Reagan to be opposites. For Strauss and Howe, Reagan and Kennedy are both members of the G.I. generation, sharing a common generational culture as "Civics." This difference in approach is made explicit by Morton Keller (1976: 134) who argues: For all the divisiveness of the New Deal, FDR and Wendell Willkie in 1940 had more in common than either did with their predecessors Smith and Hoover. In retrospect, Dwight Eisenhower and Adlai Stevenson seem to have been cut from the same consensual cloth of the 1950s. And Richard Nixon and Jimmy Carter may y et be seen as exemplars of the current American political generation: men of the New West and the New South, attaining the presidency through their mastery of a politics of media imagery that has supplanted a politics of organization, interests, and issues. DASSBACH. Carl Dassbach (1993) has explored the interesting question of the relationship between long waves in economic life and historical generations. Based on Goldstein's economic history (1988), Dassbach argues that long waves last approximately 50 years, which are split into an expansionary A-phase and a contractionary B-phase. These expansionary and contractionary periods are thus of the right length to match up with the Strauss and Howe's 22-year generations, and two of Dassbach's long waves could match up with Strauss and Howe's four generation cycle. Dassbach provides some empirical support for matching long waves with generational phenomena by correlating the long waves with Braungart's (1984) data on the frequency of youth movements over the last two centuries. The following table matches up Dassbach's dates for the economic long cycles with Strauss and Howe's dates for generational long cycles. To make the best match, I have selected the generation entering rising adulthood during each period. The historical consciousness theory of generations argues that people's consciousness is most impacted by historical conditions during the period of "youth" from approximately 16 to 25 years of age. Unfortunately, this period is not highlighted by Strauss and Howe, who split the life cycle between "youth" and "rising adulthood" at age 22. TABLE SIX Economic Economic Generational Generation Generational period condition period Entering Type Rising Adulthood 1814-1848 Contract- 1814-1843 Transcendental Idealist ion 1848-1872 Expansion 1844-1864 Gilded Reactive 1865-1881 Progressive Adaptive 1872-1893 Contract- 1882-1904 Missionary Idealist ion 1893-1918 Expansion 1905-1922 Lost Reactive 1918-1945 Contract- 1923-1946 G.I. Civic ion 1945-1968 Expansion 1947-1964 Silent Adaptive 1968-[1993]Contra- 1965-1982 Baby Boom Idealist ction [1993-2018][Expan- 1983-2003 Thirteenth Reactive sion] The match between the theories is rough as one would expect given the nature of the data. Interestingly, the biggest gap occurs in the Civil War period, which is precisely when Strauss and Howe's own theory doesn't work. The more active generations (Civic and Idealist) seem to be those who experience their formative years during a period of economic contraction. They claim that the Civil War happened too soon, and for this reason there was no Civic generation after the reactive Gilded generation. Wars are important to both the economic long-cycle theory (Goldstein, 1988) and Strauss and Howe's generational long-cycle theory (most of their "secular crises" involve wars). Very likely the wars cause the economic trends detected by long- cycle theorists, although economic determinists would draw the causal path in the opposite direction. Generational long-cycle theory would predict another major war (or equivalent "secular crisis") in the 2020's. Economic long-cycle theory would predict one about now (50 years after World War II). All in all, in my view, the generational cycle theory seems to fit the historical data better, and the economic long-cycles might be viewed as a consequence of generational cycles. So far as I know, this has not been considered in the long-cycle literature, which is dominated by neo-Marxists. LASSWELL. Harold Lasswell (1932) was perhaps the first to try to relate underlying psychodynamic factors to cycles in mass political consciousness. Lasswell applied the classical Freudian schema of "id" (biological needs), "ego" (testing of reality) and "superego" (socially acquired inhibitions) to political life. He argued that it was important to consider unconscious motivations, e.g., in understanding why leaders such as Woodrow Wilson and Abraham Lincoln behaved irrationally, or why voters tend to view political leaders as saints or martyrs. He suggested that his ideas could be applied to social dynamics, but did not develop the idea fully. He summarized his essential principle as follows (1932: 538): "prolonged ego and superego indulgence produces redefinitions in directions gratifying to the id; prolonged ego and id indulgence produces redefinitions in directions gratifying to the superego." A number of interesting questions are raised by this schema. Presumably "ego" coincides with moderate or pragmatic politics, with "id" and "superego" corresponding to the extremes. Can "id" and "superego" be equated with "left" and "right," or are both extremes characterized by excesses of id and superego? Do they correspond to the Schlesinger's eras of dominance by "public purposes" (superego) and "private interests" (id). Why should the society as a whole tend to cycle from one to the other? DEMAUSE. A much more developed theory is that of Lloyd DeMause (1982) whose theory of historical group-fantasies is an ambitious and creative psychodynamic explanation of the cycles in mass political consciousness. Group fantasies are created because people displace into the public realm feelings connected with their search for love. These collective fantasies help people to express feelings which are otherwise private and to "act out and defend against repressed desires, rages and prohibitions which have their origins in childhoods common to the group" (p. 172). Group fantasies function as defense mechanisms and are unstable just as individual defense mechanisms are. They are therefore "subject to eventual breakdown due both to the return of the repressed and to the inability of reality to live up to the requirements of the fantasy content" (p. 174). DeMause's theory is complex and varied; he immerses himself in images from the mass media and modifies his theory incorporate new insights as he goes along. One idea, developed in his study of Jimmy Carter's Presidency, is that the group's confidence in its leader goes through four regular stages, which can, but need not be, seen as metaphors for the birth process. DeMause believes that people have memories of fetal events, such as shortages of oxygen from the placenta, and of the birth process, which persist into adulthood and shape political views. He makes an interesting historical case for this argument, but one need not accept it to find value in his observations about cycles in mass consciousness. His stages are shorter than those of the Schlesingers or Klingberg, to say nothing of Sorokin. In fact, they fit well with the four years of an American Presidential term, which reflects DeMause's intense interest in day-to-day American politics. His four stages are as follows: TABLE SEVEN STRONG The first year of a leader's term. Media images stress the strength of the leader and the nation. The leader is a container for the group's fantasy of strength and safety. CRACKING The group's feeling of strength and safety weakens, there is an increase in scapegoating to avoid blaming the leader, and the group's boundaries seem to be cracking or crumbling. COLLAPSE Intense anxiety about the collapse of the group's fantasy of strength leads to rage against the leader. UPHEAVAL A "group-psychotic insight" identifies a "delusional poisoner" who is the target for the group's rage. The delusional poisoner may be the leader, who may be ritually killed, or it may be an external enemy who can be provoked and attacked. Failing these groups, the nation may become suicidal, sacrificing poisoned elements of itself to an outside enemy. DeMause sees these cycles as part of a spiral pattern of history, i.e., there is a long term progressive trend caused by generational improvements in child rearing. There are two "psychoclasses" which alternate in dominance from cycle to cycle. The first, the "liberals" is the most progressive. The liberals identify with the id. They fear separation and seek security in revolt. The "conservatives" are the less advanced psychoclass. They identify with the superego, fear gratification, and seek security in order and discipline. Thus, DeMause sees the same groups as other observers - liberals and conservatives - but he argues that this difference is based in their psychological makeup rather than in their economic interests or cognitive beliefs. DeMause's theory offers so many options that it is impossible for him to make definite predictions. His spiral metaphor would suggest that liberals and conservatives should succeed each other regularly in four year intervals, but it is obvious that they do not and DeMause does not address the reasons why not. DeMause offers interesting hypotheses about regular- ities in history, e.g., that violent group-delusions are preceded by conspiracy theories and preoccupation with sexual goings-on. DeMause also frequently offers the readers of his Journal of Psychohistory dire warnings about likely future events. For example, he predicted that "Jimmy Carter - for reasons rooted both in his own personality and in the powerful emotional demands of American fantasy - is very likely to lead us into a new war by 1979" (Demause, 1982: 147). In fact, Carter chose not to take the opportunity for war provided by the Iranian hostage crisis, preferring to allow his popularity to wane to the point where he lost the 1980 election. The weakest part of DeMause's argument is his use of evidence. He illustrates his observations profusely with selected historical references and media images, but without any rigorous attempt to determine whether the alleged patterns can be confirmed statistically. He argues that "the body images used by the cartoonists of the nation are by far the best index of the group-fantasy stages of the nation" (DeMause, 1982: 204). He illustrates his writings with cartoons which fit his themes, but has never conducted a systematic content analysis to determine whether there are, in fact, cyclical variations in the body images in published cartoons. He was enthusiastic when I (Goertzel, forthcoming) undertook such a project, but when the results did not fit his theories he became quite defensive. He objected to the way I selected cartoons for analysis, and insisted that only certain elite cartoonists could be relied upon to be truly in touch with the national mood. In his own analyses, he apparently selects out the cartoons which fit his theory and simply ignores those which do not. This is a defense mechanism frequently used by true believers (Goertzel, 1982) when evidence does not fit their preconceptions. Although my content analysis did not find the pattern which DeMause had hypothesized as leading up to the Gulf War, there was some suggestion of a cyclical pattern in the amount of emotional symbolism present in editorial cartoons, particularly with regard to the themes of indulgence, death and enemy imagery. The cycles appeared to be of approximately six to eight months duration. They need to be confirmed with more data. MAYER. Finally, we can examine a very impressive empirical work which tries to base the study of cycles in quantitative survey data. William Mayer (1992a, 1992b) offers the most conclusive evidence for the reality of cycles in public opinion in the United States. He has tabulated the results of hundreds of public opinion polls conducted over the period from 1960 to 1988. Frequently these surveys have asked the same question at several points in time. There are problems with the data such as variations in sample designs, the effects of question order, wording changes or changes in the meaning of words over time, etc. Nevertheless, this does provide a more rigorous and objective measure of changes in attitudes than more subjective interpretations of election results, newspaper cartoons or Presidential speeches. A major limitation, of course, is that he was only able to find sufficient data back to 1960, and even then the data were thin before 1975. How much difference does this quantitative rigor make? Although Mayer discusses a relatively short time period, it is one for which many more subjective interpretations are available. Some left-leaning writers, for example, have written books and articles denying that public opinion took a turn to the right in the 1970s (Parenti, 1986; Paletz and Entman, 1981). These writers claimed that the "Reagan revolution" was not in touch with public opinion, but distorted and manipulated its interpretation. Mayer's data does lead to a different conclusion than these writers. His characterization of the changes in the period are shown in the table. Mayer uses the terms "liberal" and "conservative" to organize his analysis of a great many public opinion items. Many public opinion specialists question how well the public's attitudes can actually be clustered into these two categories (Converse, 1964; Goertzel, 1981). Mayer avoids this issue by examining the survey results for each item separately, never examining the empirical validity of the liberal-conservative concept which he assumes to be valid because most political analysts use it. He detects a liberal shift when opinion moves in a liberal direction on more survey items than it moves in the conservative direction. This method gives equal weight to all issues measured by public opinion polls, while the public may actually care more about some of the issues than others. This assumes that pollsters are doing a good job of selecting the issues of most importance to the public. An alternative might be to examine questions which ask the public which issues concern them the most, or to focus on those issues which correlate most closely with voting behavior. TABLE EIGHT 1960- 1965 STABILITY. Liberal trends on racial equality and capital punishment. Most changes small and counterbalanced by conservative trends. 1966- 1973 LIBERAL SHIFT. Liberal changes on premarital sex, abortion, racial equality, women's status, foreign policy, the defense budget, Soviet relations and many other issues. Conservative trends on crime, taxes, labor, gun ownership, foreign aid. 1974- 1980 RIGHT TURN. Conservative changes on crime, the ERA, Soviet relations, military spending and intervention, taxes, government spending, regulations, inflation and the environment. Liberal trends on racial equality, women, premarital sex, legalizing marijuana. 1981- 1988 LIBERAL RESURGENCE. Liberal changes on race, military, government spending, gay rights, Soviet relations, women, family, environment. Conservative trends on crime, abortion, economic rights & privileges, business profits. Mayer's detailed issue-specific data makes it possible to ask whether there are different trends for different kinds of issues, e.g., foreign policy issues, economic issues, life style issues, etc. Some of these may be cyclical, others may not. In fact, Mayer finds that attitude change is even more issue- specific than that. For example, Americans have become more liberal on women's rights, but more conservative on the Equal Rights Ammendment. They have become more conservative on gun ownership, but not on gun registration. There has been a strong liberal trend on race relations and women's equality, and a consistent conservative trend on crime and punishment. The survey data allow Mayer to look at age differences, providing an empirical test of the generational hypotheses which Strauss and Howe base on cultural interpretation. His results are mixed. Generational effects have been important for changes on social and cultural issues, but not for opinions about foreign policy and economic issues. Perhaps because of his immersion in this mass of data, Mayer has a hard time reaching conclusions about master trends. He observed that public opinion had been liberalizing from 1981- 1988, but he failed to predict that this trend would have an effect on the 1992 Presidential elections. Succumbing to the conventional wisdom of his time, he discounted his own data and predicted "the most likely scenario for the future is a replay of the last ten years; more Republican presidents, and more liberal anguish and breast-beating" (Mayer, 1992b: 340). COVER. Another attempt to examine cyclical trends empirically is that of Daniel Cover (1992) who has content analyzed Time Magazine covers from 1923 to 1988 looking for cycles in the kinds of social roles portrayed. He categorized these roles according to Talcott Parson's A-G-I-L organizational cycle theory as adapted by Suzanne Keller (1991: 260). Leaders were classified into "internal" and "external" categories. External leaders were those in business, politics, science, government or the military whose functions were primarily adaptation or goal attainment (A and G in Parson's schema). The internal leaders were movie stars, entertainers and media personalities, athletes, writers, artists, labor leaders, first family members, philosophers, educators, and clergy whose functions were integration or latency (I and L in Parson's schema). He found that the percent of external roles portrayed on Time covers was moderately (r=+.41) correlated with the percent of the vote for Republican presidential candidates. His work is in process, and more data are needed to verify the trends. A visual inspection of his time series graphs suggests that the cycles in the percentage of external roles portrayed are short, perhaps six to eight years. Cover's work is interesting because it provides additional empirical support for the reality of cycles, and because it suggests an additional theoretical interpretation. CONCLUSIONS. The phenomenon of cycles is very complex, involving a number of dimensions which are only loosely related. The data are very sketchy, allowing for alternative interpretations as analysts from different schools of thought focus on the phenomena which fit their models. Some tentative observations can be offered, but only as a guide to further discussion and research. Trends in generational culture seem to have little to do with the political alternation between liberals and conservatives. These are perhaps best studied separately. The Schlesinger's model of approximately 15 year liberal- conservative cycles is about as well supported as anything in this field, and seems to be related to organizational dynamics as much as to public opinion. It seems to take about that long for a political movement to work through the process of putting together a successful political coalition, institutionalizing its policies, and disillusioning the public with its failure to accomplish its goals. Although many analysts fall back on the liberal-conservative concept, the empirical data suggests that this is very crude and obscures many important differences. Opinion on social issues often varies quite independently from that on economic issues. Foreign policy issues may vary independently from domestic policy. There is some evidence for the existence of generational long-cycles, especially on cultural issues, and these may be related to economic long-cycles. The fact that biological generations are continuous, however, makes it difficult to clearly define historical generations. Work on cycles in the mass unconscious is in its infancy, mostly the provocative and lavishly illustrated speculations of DeMause and his followers. It may be that the lack of adequate data is simply too large a barrier to overcome. Psychobiographical investigation might offer a useful approach here. Studies of the psychobiography of political candidates from different eras might offer insights into changes in the mass psychology. Generational theory could offer a useful approach here, since it provides a way of uniting family and historical dynamics, but the issue has not been adequately pursued. DeMause's work emphasizes a much shorter time cycle, and may capture a pattern which is built into the four year Presidential election cycle. Just as in economics, there may be both short and long term cycles in political psychology. Untangling these is very difficult perhaps because of the paucity of hard data. However, the topic also remains confused in economics, the complexity of a system with overlapping periodic cycles may lead to inherently chaotic outcomes. Certainly the record of prediction by cyclical analysts is spotty, but they do draw attention to turning points which are missed entirely by other theoretical approaches. The best results seem to be those of the Schlesingers, which focus on organizational dynamics in the political parties. REFERENCES Braungart, Richard. (1984). "Historical Generations and Generation Units: A Global Pattern of Youth Movements." Journal of Political and Military Sociology 12: 113-135. Campbell, Philip. (1964). "The nature of belief systems in mass publics." Pp. 206-221 in David Apter, ed., Ideology and Discontent. New York: Free Press. Cover, J. Daniel. (1992). "Cycles of political conservatism from 1924-1988." Paper presented at the annual meetings of the American Sociological Association. (Sociology Dept., Furman University, Greenville SC 29613). Dassbach, Carl. (1993). "Long Waves and Historical Generations: A World Systems Approach," paper prepared for the 1993 ISPP meetings. DeMause, Lloyd. (1982). Foundations of Psychohistory. New York: Creative Roots. Elazar, Daniel. (1976). The Generational Rhythm of American Politics. Philadelphia: Center for the Study of Federalism, Temple University. Ferrari, Giuseppe. (1872). Teoria dei Periodi Politici. Goertzel, Ted. (1982). Turncoats and True Believers. Buffalo NY: Prometheus. Goertzel, Ted. (1972). "Generational Conflict and Social Change," Youth and Society 3: 327-352. Goertzel, Ted. (forthcoming). "The Gulf War as a Mental Disorder: A Statistical Test of DeMause's Hypothesis." Political Psychology. Goertzel, Ted. (1981). "Ideological consistency and issue voting in 1980." Journal of Political and Military Sociology 9: 197-214. Goldstein, J. (1988). Long Cycles: Prosperity and War in the Modern Age. New Haven: Yale University Press. Huntington, Samuel. (1974). "Paradigms of American Politics: Beyond the One, the Two and the Many." Political Science Quarterly, May. Huntington, Samuel. (1981). American Politics: The Promise of Disharmony, p. 147. Keller, Suzanne. (1991). Beyond the Ruling Class. New Brunswick, NJ: Transaction Books Keller, Morton. (1979). "Reflections on Politics and Generations in America." Pp. 123-135 in Stephen Graubard, ed., Generations, New York: Norton, 1979. Klinberg, Frank. (1952). The historical alternation of moods in American foreign policy. World Politics 4: 239-273. Littre, Emile. (1860). Paroles de philosophie positive. Mannheim, Karl. (1952). "The Problem of Generations." Pp. 276- 322 in his Essays on the Sociology of Knowledge. London: Routledge and Kegan Paul, 1952. Marias, Julian. (1968). "Generations" in the International Encyclopedia of the Social Sciences. Marias, Julian. (1967). Generations: A Historical Method. University of Alabama Press translation, 1970. Mayer, William. (1992b). The Changing American Mind: How and Why American Public Opinion Changes between 1960 and 1988. Ann Arbor: University of Michigan Press. Mayer, William. (1992a). "The shifting sands of public opinion: Is liberalism back?" The Public Interest 107: 3-17. Paletz, David and Entman, Robert. (1981). Media Power Politics. New York: Free Press. Parenti, Michael. (1986). Inventing Reality: The Politics of the Mass Media. New York: St. Martin's. Schlesinger, Arthur M., Jr. (1986). The Cycles of American History. Boston: Houghton Mifflin. Schlesinger, Arthur M., Jr. (1992). The turn of the cycle. The New Yorker, 68 (November 16), 46-54. Schlesinger, Arthur M., Jr. (1972). "How McGovern will win." New York Times Magazine, July 30. Schlesinger, Arthur M., Sr. (1949) Paths to the Present. New York: MacMillan. Simonton, Dean Keith. (1984). Genius, Creativity & Leadership. Cambridge: Harvard University Press. Simonton, Dean Keith. (1976). Does Sorokin's data support his theory?: A study of generational fluctuations in philosophical beliefs. Journal for the Scientifric Study of Religion 15: 187-198. Sorokin, Pitirim. (1941). Social and Cultural Dynamics, Volume Four: Basic Problems, Principles and Methods, New York: American Book Company. Sorokin, Pitirim. (1957). Social and Cultural Dynamics, One Volume Revision. Boston: Porter Sargent. Strauss, William and Neil Howe. (1991). Generations: The History of America's Future, 1584 to 2069. New York: William Morrow and Company.