Boeing Co. officially launched its 7E7 Dreamliner passenger jet Monday after Japan’s All Nippon Airways Co. placed the first order for the fuel-efficient, dolphin-shaped plane that Boeing hopes will help it regain supremacy in the commercial aviation market.

The order for 50 jetliners is the largest by a single airline in the history of Boeing, which valued it at $6 billion based on 7E7 price lists.

“It’s a historic moment for us,” said Mike Bair, Boeing’s senior vice president for the 7E7 program.

Los Angeles Times, April 27, 2004

Anyway . . . I agree with Rick Perlstein that the Democrats should emphasize a message of economic populism. I am not sure that such a strategy is as beneficial as Perlstein suggests, but that does not matter to me. Economic populism is morally right regardless of whether it gains votes.

I do, however, disagree with Perlstein about his analysis of the Democratic Party’s current state and future prospects. The concept of party identification underpins Perlstein’s analysis, since party ID “is the most reliable predictor of whether someone will vote for a given candidate. It is a mighty store of value, party identity.” Consequently, he cites recent declines in the Democratic advantage in party ID as evidence of the Democrats’ current problems and sees expanding party ID as central to the Democrats’ future prospects.

Unfortunately, there’s no clear evidence of a significant decline in the Democrats’ party ID in recent decades. According to Perlstein, Democratic identifiers have fallen from a high of 51 percent in 1977 to about 45 percent today. Thus, the Democrats have lost about 5 percentage points over the last 25 years. That’s a decline, but hardly a major one.

Though they have not lost much ground in an absolute sense, Democrats have, as Perlstein points out, lost ground compared to the Republicans. This relative decline resulted largely because the percentage of Republican identifiers has increased by about 15 percentage points since the late 1970s. It therefore bears asking why the electorate has become more Republican in recent decades. Perlstein’s analysis leads one to conclude that otherwise economically liberal voters shifted to the Republicans because they perceived the Democratic party as too wishy-washy or cravenly pragmatic.

But an analysis of the data tells a very different story. First, as Table 1 shows, the American public has become more conservative since the late 1970s. Second, as shown in Table 2, since that time, conservatives have shifted heavily toward the Republicans, more than offsetting Democratic gains among liberals. Thus, despite Perlstein’s analysis, recent Republican gains have had little if anything to do with the Democrats abandoning their economic liberalism. Instead, the American public has become more conservative and those new conservatives were more likely to affiliate with the Republican Party.

Perlstein is right when he says that party identification, more than any other factor, influences how people vote, but this does not, as he implies, translate into any consistent electoral advantage. Indeed, the Republicans, despite their disadvantage in party identification over the last 50 years, have still managed to win a majority (seven out of 12) of the presidential elections in that period. Why is this? Not all party identifiers vote, and of those that do, not all of them vote for their party’s candidates. Furthermore, though independent voters only make up about 10 percent of the electorate, that is more than enough to swing most elections. As a result, a strategy of winning elections by relying solely on the votes of one’s partisans has never been enough for victory. To a greater or lesser degree, any party that hopes to win an election will have to cater to the “stock-ticker rules” and “short-term whims” that Perlstein bemoans.

But even if Perlstein were right about party identification, I would still be skeptical about the utility of his proposal for Democrats to rethink themselves. Since George Washington ran for reelection in 1792, there have been 53 presidential elections. (I know that Perlstein doesn’t focus just on winning presidential elections, but they are the most important and consequential elections for any party.) Twenty-nine of those elections involved incumbents running for reelection, of which the incumbent won 20, or 69 percent. The other 24 elections were open races with no incumbent, and in half of them, the party out of power won the election. In other words, if you take incumbents out of the picture, a party has a 50–50 shot at winning the next presidential election. I take this to mean that winning elections is a random sort of thing and not something that parties have much control over. Any sort of effort to rethink the Democratic Party and what it stands for is a rather futile proposition because events will outrun even the most perspicacious observers. Elections tend to turn on such unpredictable and uncontrollable events as wars, recessions, and scandals. Indeed, but for a semen-stained dress and confusing ballots in Palm Beach County, it would be the Republicans and not the Democrats who would be undergoing this bout of introspection, self-criticism, and self-flagellation.

Furthermore, if elections are hard to influence, party identification is even worse. As Perlstein mentions not once but twice in his article, party identification is a stable and long-lasting aspect of social identity. As such, it is difficult if not impossible for a party to influence party identification absent some major event—such as the Great Depression, the Cold War, or the civil-rights movement—that causes voters to reassess their partisanship.


Table 1.
Political Ideology, 1978–2002

1978 2002 Change
Liberals 26% 27% +1
Moderates 37 27 -10
Conservatives 37 45 +8

Source: Data compiled by the author from the National Election Study 1952–2002 Cumulative File.

Table 2.
Change in Party Identification by Ideology, 1978–2002

Republicans Independents Democrats
All +14 -7 -7
Liberals +1 -8 +8
Moderates +9 -7 -2
Conservatives +21 -7 -14

Source: Data compiled by the author from the National Election Study 1952–2002 Cumulative File.

Click here to return to the New Democracy Forum “How Can the Democrats Win?

Originally published in the summer 2004 issue of Boston Review.