This is section seven

JOURNAL THE POLLSTER GEIST

Ronald Gordon Ziegler I. The Perot Effect Any number of accounts of the 1996 elections will be entered, most making up in enthusiasm and creativity whatever they may lack in substantive interpretation. Too many pundits, political scientists, press, and Democrat politicians are all echoing variations of the same theme. As they have increasingly since 1972 (and before), they are marching in lockstep. They are at present proclaiming a decisive electoral victory for Clinton. In reality, whatever it may have been, that was not it. There are some important reasons that Dole's effort fell short. Understanding these offers some insights into and criticisms of our political system which must not be ignored. There is one big reason that neither Bush in his re-election bid nor Dole in his drive to oust Clinton were successful. Or, conversely, there is that one big reason that Clinton was able to 'win' the two contests. His name is Ross Perot. If you remove him from the picture, both Bush and Dole win. But there were also media effects and informational cost effects which played important roles in these two races. As for the Perot Effect, it can be seen clearly in the popular vote for both elections. It is no less a certitude in the Electoral College.

Table I -- Voting in 1992 and 1996(NYT, 11/7/96) 1992 1996 Clinton 43% 370 49% 379 Republican 38% 168 41% 159 Perot 19% 10%

There is an illusion of perhaps a 'strong' victory, but it is only illusion. The margin of Clinton's victory was so thin in so many states that the Perot voter is probably a singular difference in the outcome. In total popular votes, the Republican and Democrat candidates collected roughly the same amount of votes in 1996 as the standard bearers had received in 1992 ( NYT 11/7/96). The big apparent change in percentage of the vote has to do with the decline in voter turn-out and in the Perot vote. Some have speculated that the divided government result of the election is a sort of collective conscious decision made by the electorate. But it is not, as so widely touted, so much indicative of ticket splitting in any traditional sense, although that clearly occurred. It gives added strength to the notion that the vote was a rejection of Clinton by a majority of voters, but one divided between two 'alternatives.' But as important as what the vote was 'against' is what the vote was 'for.' Otherwise, a similar argument could be constructed for 1992 as a rejection of Bush. It was effectively that, of course, but it was also more complicated than that. A consideration of GOP vote erosion in state after state is clearly indicative of a siphoning off of Republican base vote at the Presidential level in states which had solid Republican majorities and which continued to do so at levels other than the Presidential vote. Nor can it be overlooked that 1996 represents a quite peculiar phenomenon in American presidential politics. When was the last time a President seeking re-election got less than half of the vote? It has happened, of course . . . to Bush, Carter, Hoover, and Taft in this century. But what do they have in common? They all were defeated. When was the last time a President seeking re-election got less than half the votes and won? It ain't happened! (Truman was elected in 1948 with less than fifty percent of the vote, but he was being 'elected,' having become President when FDR died). II. Vote Deficit What is of consequence is that the campaigns were effected and affected. But what seems to have been overlooked has been this impact of Perot's races on the election, whatever the motivation behind it, for the substantive results they created. Political scientists and pundits alike seem to be unable to acknowledge that one fact of political life. But a strong argument can be offered that Perot's presence in the races adversely affected both the Bush and Dole campaigns causing their loss. It is quite likely -- indeed, an almost certainty -- that Clinton could not have won either time without that factor. A consideration of returns across the nation fuels that analysis. In seventeen states, with 193 electoral votes, there was, in comparing 1992 with other recent elections, a real reason to believe that if Perot had not been in the race, Bush would have carried them. In few of these states was there a decline in the Democrat vote. However, for the Republican ticket, the drop in each one was of significant proportion. Tracking the vote in each over many elections seems to make 1992 aberrational. In a number of other states, such as Indiana, the Bush vote seems to have been significantly reduced. Had Perot never gotten into the race, or if he had even remained out once he had withdrawn, it seems altogether plausible that not only would Bush have won re-election with perhaps 361 electoral votes, but that he would have readily carried the popular vote, as well. Bush would not have needed all of these. He won 188 votes in the Electoral College, and there were enough of these seventeen states where the vote was quite close enough that he could have gotten the requisite 270 but for Perot. Even without California, Connecticut, Iowa, Michigan, and Vermont, Bush would have won 271. And given both other races in these states and subsequent electoral returns in them, more weight is lent to the possibility.

Table II -- 1992 and 1988 Presidential Votes (Voter News Service in World Almanac 1996 ) 1992 1988 Electors State Clinton Bush Perot Dukakis Bush 54 Calif 4812317 3338942 2144856 4702233 5054917 8 Colo 625402 555408 362506 621453 728177 8 Conn 681081 574738 348028 676584 750241 3 Dela 125997 102436 59061 108647 139639 13 Ga 1002433 985682 306489 714792 1081331 7 Iowa 583937 503338 251040 670557 545355 8 Kent 661059 617419 203587 580368 734281 9 Louisiana 815305 729880 210604 717460 883702 4 Maine 263420 206504 206820 243569 307131 18 Michigan 1854603 1585251 819931 1675783 1965486 3 Montana 153899 143702 106735 168936 190412 4 Nevada 185401 171378 129532 132738 206040 4 N.H. 207264 199623 120029 163696 281537 15 N.J. 1361842 1303686 504152 1317541 1740604 21 Ohio 1964942 1876445 1024319 1930629 2416549 3 Vermont 125803 85512 61510 115775 124331 11 Wisconsin 1035942 926245 542610 1126794 1047499 ___ 193

What can be seen in looking at voting returns in the last eleven Presidential elections in each of these states is the clear indication of the 'vote deficit' Bush suffered a loss in to Perot.

Table III -- California Presidential Voting (Voter News Service) Year Republican Democrat Third 1952 2.89 2.20 1956 3.03 2.42 1960 3.26 3.22 1964 2.88 4.17 1968 3.47 3.24 .5 1972 4.60 3.48 1976 3.88 3.74 1980 4.52 3.08 .7 1984 5.31 3.82 1988 5.05 4.70 1992 3.34 4.81 2.14 1996 3.42 4.64 .67

Variations in each party's vote in each of the past eleven Presidential contests are around a line sloping upwards which could be suggested as a 'predictor' of expected or potential vote for each party. (The slope could conceivably be downward, kinked, or even level, but general population growth trends in most states should create the probability of an upward slope in most). The variations seem to reflect short term impacts in the long term trend. On this analysis (see Appendix A), Bush should have been expected to carry California by about 600000 votes, having won perhaps as much as 5.5 million votes to Clinton's 4.9 million. The 'deficit' in this number compared with the actual 1992 vote is 2.2 million for the Republicans and about .2 million for the Democrats. It is argued that the character of the three way race increased voter turn-out somewhat, but it would seem that most of the 'lost' Bush vote went to Perot. If voters in California had followed the vote growth pattern of the most recent elections, the 'average' gain might have been .24 for the GOP, giving them 5.29, and .18 for the Democrats, setting their tally slightly higher than it was at 4.88. 'Average' can be a very misleading indicator, however, especially in this type of contest. The pattern for both parties seems to have been a period of growth followed by a 'correction.' Although the low growth of the Democrat vote could be seen as a 'settling,' there is little to indicate anything normal about the GOP decline. Whether it should have been expected to hit 5.29 (or 5.5), or more or slightly less, it should not have dropped much at all, let alone significantly, especially against the backdrop of growing Republican strength in the state. The only explanation for the big drop has to be Perot. Even the 'bad' Republican years of 1964 and 1976 pale against this drop. In percent change, the decline represents 1/3 of the GOP vote in '88 (which was down itself from '84) compared to the mere 12% drop in 1964, and 16% in 1976. Similar patterns can be plotted for each of the states mentioned (again, see Appendix A), as well as for the nation as a whole. A synopsis of projected vote and vote deficit for each of the states mentioned can be constructed for 1992, indicative of the impact which Perot had on drawing off votes which would otherwise have gone to Bush.

Table IV -- National Presidential Voting (Voting News Service 1996) Year Republican Democrat Third 1952 33.9m 27.3m 1956 35.6 26.0 1960 34.1 34.2 1964 27.2 43.8 1968 31.8 31.3 9.9 1972 41.2 29.2 1976 39.2 40.8 1980 43.9 35.5 5.7 1984 54.3 37.5 1988 48.9 41.8 1992 39.1 44.9 19.7 1996 37.9 45.6 7.9 [2000] [50] [47] Continue