² Quoted, Wise, David, “The Twilight of a President,” New York Times Magazine, 11 3, 1968, p. 131Google Scholar.

³ Neustadt, Richard E., Presidential Power: The Politics of Leadership (New York: Wiley, 1960), p. 205nGoogle Scholar.

⁴ A general picture of what this line looks like can be gained from the figure in Dahl, Robert A., Pluralist Democracy in the United States (Chicago: Rand McNally, 1967), p. 107Google Scholar. The Presidential popularity data for the Johnson administration have been taken from the Gallup Opinion Index. All other poll data, unless otherwise indicated, have come from the archives of the Roper Public Opinion Research Center at Williams College, Williamstown, Massachusetts.

⁵ Op. cit., p. 96.

⁶ See for example the breakdowns in Gallup Opinion Index, 03, 1966, p. 4Google Scholar.

⁷ There is also a more technical reason why the popularity index has little direct relevance to the electoral result: Gallup does not ask the question during a President's re-election campaign. Thus for the months between early summer and late fall in 1948, 1956, and 1964 no Gallup data on Presidential popularity exist. One other technicality is worth mention. There ia a slight underrepresentation of data points in Truman's first years. By 1950, except for the election year phenomenon, the Gallup organization was asking the question on virtually every survey conducted—some dozen or sixteen per year. Before that time the question was posed on the average only about half as frequently. Neither of these technical problems, however, is likely to bias the results in any important way, especially since so much of the analysis allows each administration a fair amount of distinctiveness.

⁸ The standard deviation for the no opinion response is 2.93. By contrast the comparable statistic is 14.8 for the approve response and 14.5 for the disapprove response.

⁹ In the case of President Eisenhower, the no opinion response actually rose a bit before it began to descend, reaching the highest level recorded for any President in the period in March 1953 when 28 percent had no opinion. (President Nixon has proved to be the greatest mystery of all: fully 36 percent registered no opinion after his inauguration.)

¹⁰ It is argued by some that percentages should not be used in their pure state as variables, but rather should be transformed into logits: Y* = log_e [Y/(1−Y)]. The transformation was tried in the analysis, but it made little difference. Therefore the more easily communicated percentage version haa been kept. In any event the dependent variable rarely takes extreme values. It rises to 80 percent only three or four times and never dips below 23 percent.

¹¹ An Economic Theory of Democracy (New York: Harper and Row, 1957), pp. 55–60Google Scholar.

¹² See the data in Erskine, H. G., “The Polls,” 28 Public Opinion Quarterly, 341 and 338 (Summer 1964)Google Scholar.

¹³ See Fenton, John M., In Your Opinion (Boston: Little, Brown, 1960), p. 146Google Scholar.

¹⁴ “The Empty-head Blues: Black Rebellion and White Reaction.” The Public Interest, Spring 1968, pp. 3–16Google Scholar.

¹⁵ On the bandwagon effect among nonvoters, see Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., The American Voter (New York: Wiley, 1960), pp. 110–115Google Scholar.

¹⁶ Roper, Burns, “The Public Looks at Presidents,” The Public Pulse, 01 1969Google Scholar.

¹⁷ The Prince, ch. XVII.

¹⁸ Waltz, Kenneth N., “Electoral Punishment and Foreign Policy Crisis” in Rosenau, James N., Domestic Sources of Foreign Policy (New York: Free Press, 1967), p. 272Google Scholar.

¹⁹ Wicker, Tom, “In the Nation: Peace, It's Wonderful,” New York Times, 07 4, 1967, p. 18Google Scholar.

²⁰ Op. cit., p. 100.

²¹ Op. cit.

²² Congress and the Presidency (Englewood Cliffs, N. J.: Prentice-Hall, 1964), p. 25Google Scholar.

²³ Especially valuable was Nordheim, Eric V. and Wilcox, Pamela B., “Major Events of the Nuclear Age: A Chronology to Assist in the Analysis of American Public Opinion,” Oak Ridge National Laboratory, Oak Ridge, Tennessee, 08 1967CrossRefGoogle Scholar. Other sources often consulted included the New York Times Index and the Chronology section of the World Almanac.

²⁴ Roper notes that President Kennedy's highest point of popularity occurred after the Bay of Pigs invasion and concludes this fact says something special about that crisis event (op. cit.) But this phenomenon is due to two effects—the rally round the flag effect and the fact that the event occurred very early in Kennedy's administration when the value for the coalition of minorities variable was yet very low.

²⁵ Op. cit., p. 97ff.

²⁶ Data were gathered from Moore, Geoffrey H. (ed.), Business Cycle Indicators, Vol. II (Princeton: Princeton University Press, 1961), p. 122Google Scholar; and, for more recent data, issues of the Monthly Labor Review.

²⁷ One wrinkle, which is intuitively comfortable but makes little difference in the actual results, was to do something about the unemployment rates at the start of the first terms of Presidents Truman and Eisenhower when unemployment was “artifically” depressed due to ongoing wars. Presumably the public would be understanding about the immediate postwar rise in unemployment. Therefore for these two terms the initial unemployment level was taken to be that level which held six months after the war ended while the economic variable for the few months of the war and the six month period was set equal to zero.

²⁸ “The Pain Threshold of Economics in an Election Year,” New York Times, 07 15, 1968Google Scholar.

²⁹ Waltz, op. cit., pp. 273ff, 288; Neustadt, op. cit., pp. 97–99; Wicker, op. cit.; Roper, op. cit.

³⁰ Steele, A. T., The American People and China (New York: McGraw-Hill, 1966), p. 294Google Scholar. In May 1964 Gallup found almost two-thirds of the population said they paid little or no attention to developments in South Vietnam. Free, Lloyd A. and Cantril, Hadley, The Political Beliefs of Americans (New York: Clarion, 1968), pp. 59–60Google Scholar.

³¹ Although the Korean War continued into President Eisenhower's administration, he is not “blamed” for the war in the analysis since of course he was elected partly because of discontent over the war. Accordingly the war variable is set at zero for this period.

³² Each equation is displayed vertically. The dependent variable, the percentage approving the way the President is handling his job, has a mean of 57.5 and a standard deviation of 14.8. The number of cases is 292. The figures in parentheses are the standard errors for the respective partial regression coefficients. To be regarded statistically significant a regression coefficient should be, conventionally, at least twice its standard error. All equations reported in this study are significant (F test) at well beyond the .01 level. The Durbin-Watson d is an indicator of serial correlation which suggests decreasing positive serial correlation as the statistic approaches the value of 2.0. All equations in this study exhibit a statistically significant amount of positive serial correlation.

³³ The dummy variables formalize the sort of discussion found in Neustadt, op. cit., p. 98. They account for what a singer might call tessitura.

³⁴ It was noted in Section I that some minor bias in these results is introduced by an embellished rate of “no opinion” in the first weeks of the Kennedy and first Eisenhower terms. As this rate declined, there was some tendency for the Presidents' approval and disapproval rates to rise. To see if this peculiarity had any major impact, equations 3 and 4 were recalculated using the percentage disapproving as the dependent variable. This manipulation causes no fundamental differences, although President Eisenhower's rating behaves a little less outrageously.

³⁵ If the term dummies are dropped from the equation to attain a version comparable to equation 4, the Eisenhower phenomenon holds except that his first term increase drops to 2.00 (still significant) and his second term decrease is a slightly steeper −0.36 (still not significant).

³⁶ Regression statistics relating to President Johnson's first term are very unreliable, as the size of the standard errors suggests, because the popularity question was posed so few times during this brief period.

³⁷ Mueller, John E., “Patterns of Popular Support for the Wars in Korea and Vietnam,” unpublished paper, Department of Political Science, University of Rochester, 1969Google Scholar.

³⁸ Op. cit., p. 97.

³⁹ See Spanier, John W., The Truman-MacArthur Controversy and the Korean War (Cambridge, Mass.: Belknap, 1959)CrossRefGoogle Scholar; also Neustadt, op. cit., passim; and Higgins, Trumbull, Korea and the Fall of MacArthur (New York: Oxford, 1960)Google Scholar. See also the data in Belknap, George and Campbell, Angus, “Political Party Identification and Attitudes Toward Foreign Policy,” 15 Public Opinion Quarterly 601–23 (Winter 1951–1952)CrossRefGoogle Scholar. Note especially the strong party polarization on the issue. That the public was strongly inclined to support General MacArthur in the dispute can be seen from poll data. The first polls, conducted as the General was making his triumphal, “old soldiers never die” return to the United States in April 1951, suggest more than twice as many people supported the General as supported the President. As Neustadt suggests (op. cit., p. 97), emotion on the issue faded during the Senate Hearings on the issue which lasted until June and this seems to have been to the benefit of President Truman's position. The Truman point of view received its greatest support in late June and early July as peace talks were being begun. As the talks began to prove unproductive, however, public opinion began to revert to its previous support of General MacArthur, until, by the first days of 1952 (when the polling agencies grew bored with the issue), the MacArthur position was as strongly approved and President Truman's as strongly rejected as ever.

⁴⁰ There is evidence which suggests that World War II, a much more popular (and much larger) war than either Korea or Vietnam, may have worked to the distinct benefit of President Roosevelt. The National Opinion Research Center in its post-election poll in 1944 asked Roosevelt supporters if they would have voted for Dewey “if the war had been over.” Enough answered in the affirmative to suggest that Dewey might well have won in a warless atmosphere. From data supplied by the Inter-University Consortium for Political Research.

⁴¹ Stouffer, Samuel A., Communism, Conformity, and Civil Liberties (Garden City, N.Y.: Doubleday, 1955)Google Scholar, especially ch. 3; Campbell, et al., op. cit., pp. 50–51; Polsby, Nelson W., “Toward an Explanation of McCarthyism,” 8 Political Studies 250–71 (10 1960)CrossRefGoogle Scholar; Roper, Elmo, You and Your Leaders (New York: Morrow, 1957), pp. 250–51Google Scholar.

⁴² In late November 1945, over six months after President Roosevelt's death, Gallup asked his sample, “In your opinion, who is the greatest person living or dead, in world history?” Fully 28 percent proffered Roosevelt's name. Abraham Lincoln was mentioned by 19 percent, Jesus Christ by 15 percent, and George Washington by 8 percent. No one else received more than 2 percent. And the aura lasted. A survey conducted in June 1949 in the city of Philadelphia (which had voted 59 percent for Roosevelt in 1944 as against a national rate of 55 percent) posed this question: “Could you tell us the name of a great person, living or dead, whom you admire the most?” The most commonly mentioned names were Roosevelt with 42 percent, Lincoln with 9 percent, and Washington with 5 percent. (The absence of Jesus Christ on this latter list presumably can be laid to the peculiarities of question wording—or of Philadelphians.) Sanford, Fillmore H., “Public Orientation to Roosevelt,” 15 Public Opinion Quarterly 190–91, 200 (Summer 1951)CrossRefGoogle Scholar. In 1948, Roper found 43 percent of a national sample offering Roosevelt's name when queried, “Considering all the men in America who have been prominent in public affairs during the past 50 years, which one or two have you admired the most?” Dwight Eisenhower was second at 17 percent. E. Roper, op. cit., p. 22.

⁴³ Op. cit., p. 198.

⁴⁴ See Converse, Philip E. and Dupeux, Georges, “De Gaulle and Eisenhower: The Public Image of the Victorious General,” in Campbell, Angus, et al., Elections and the Political Order (New York: Wiley, 1966), pp. 292–345Google Scholar.

⁴⁵ In December 1960 the public was asked what it felt was Eisenhower's greatest accomplishment. The ending of the Korean War was mentioned by 11 percent and a related comment, “he kept us out of war,” was suggested by an additional 32 percent. No other specific accomplishment was mentioned by more than 5 percent; only 3 percent mentioned anything having to do with the domestic scene. See also Neustadt, op. cit., p. 98.

⁴⁶ As Irving Kristol argues, “… when a conservative administration does take office, it pursues no coherent program but merely takes satisfaction in not doing the things that the liberals may be clamoring for. This, in effect, is what happened during the two terms of President Eisenhower …” “The Old Politics, the New Politics, and the New, New Politics,” New York Times Magazine, 11 24, 1968, p. 167Google Scholar.

⁴⁷ See Rossiter, Clinton, The American Presidency (New York: Harcourt, Brace, 1960), pp. 161–78Google Scholar.

⁴⁸ Greenfield, Meg, “The Great American Morality Play,” The Reporter, 06 8, 1961, pp. 13–18Google Scholar.

⁴⁹ It may, or then again may not, be worth noting that Presidential elections in which the incumbent party was removed, 1952, 1960, and 1968, occurred during moral crises while the elections in which the President was retained, 1948, 1956, and 1964, all took place during times of relative goodness.

⁵⁰ It may prove valuable to attempt to see how spectacular and cumulative international events and shifts in governmental policy—to use the distinction made by Karl Doutsch and Richard Merritt—differ in impact. “Effects of Events on National and International Images” in Kelman, Herbert C. (ed.) International Behavior (New York: Holt, 1965), pp. 132–87Google Scholar.

⁵¹ Wesley C. Clark has found some relation between the Roosevelt popularity and the state of the economy in the 1937–1940 period. He also notes a general “downward slant” in the rating over time and finds a rise of popularity during international crises. (“Economic Aspects of a President's Popularity,” Ph.D. Dissertation, University of Pennsylvania, 1943, pp. 41, 28, 35). See also B. Roper, op. cit., and E. Roper, op. cit., chapters 2 and 3. And V. O. Key has observed that during 1940 “the popularity of Roosevelt rose and fell with European crises.” Politics, Parties and Pressure Groups (New York: Crowell, 1952, 3rd. ed.), p. 596Google Scholar (cited in Waltz, op. cit., p. 272.)

⁵² British observers have noted an apparent relation between unemployment and party preference in their country: rising unemployment seems to have benefited Labor while declining unemployment favors the Tories. Durant, Henry, “Indirect Influences on Voting Behavior,” 1 Polls 7–11 (Spring 1965)Google Scholar. Extensive data from France on the popularity of President De Gaulle have been published: for example, Gallup Opinion Index, 03 1968, pp. 27–28Google Scholar. A study by Howard Rosenthai has investigated regional aspects of the General's popularity. “The Popularity of Charles De Gaulle: Findings from Archive-based Research,” 31 Public Opinion Quarterly 381–98 (Fall 1967)CrossRefGoogle Scholar.