| This is the first comprehensive, quantitative study on the effect of George H. W. Bush's reversal of his “read my lips: no new taxes” campaign promise on the outcome of the 1992 presidential election. A multinomial unified model of vote choice and turnout is employed using 1992 National Election Studies (NE S) data. I hypothesize that the most conservative and partisan Republicans, which I define as Bush's base, either defected to another candidate or abstained from voting in protest of Bush's decision to raise taxes in 1990. Spatial voting theory would predict that an alienated base would most likely either abstain or reluctantly vote for Bush in 1992. Stephen Skowronek's (1997) discourse on political betrayal would also support an abstention thesis. Though the qualitative record shows that many conservatives openly rebelled against the president, on Election Day 1992, Bush's base supported the president more so than any other ideological-partisan group. Bush lost support, however, among non-base Republicans, independents and “Reagan Democrats.” The quantitative results also suggest that Bush lost what he hoped would be a valence advantage over Clinton on the “trust” issue, given that Bush had his own issues with trust for breaking his “read my lips” tax pledge. Had Bush stood firm on his campaign promise, he still would not have won the election, though he would have cut into Clinton's margin of victory by at least 15% and as much as 55%, depending upon the initial conditions of probability analysis. Even had Bush eliminated public misperceptions about the state of economic conditions in 1992, he still would not have changed the election outcome, though he would have reduced Clinton's margin of victory between 20% and 41%. These findings challenge the conventional wisdom, that the state of the economy (or perceptions of it) exclusively explains Bush's loss in 1992. Had Bush done both—not raised taxes and changed public misperceptions about the state of economic conditions—he would have narrowed Clinton's margin of victory between 40% and 100%. That is, in one probability analysis most favorable to Bush, he would have tied Clinton in the 1992 presidential election. |
