On Oct 21, 2008, at 6:22 PM, Charles A. Grimes wrote:
> There is something seriously wrong with this study. The only economic
> positions the Republicans offer is lower taxes, and that's a just a
> lie.
I know this is the Internet, where it's sometimes hard to read a whole paper, so I'll just quote some more from Bartels. Reading these papers, and reading Jeffrey Stonecash's book, has forced me to rethink a lot of things I used to believe. E.g., 1) the Dems would do better by taking a more "populist" economic position, and 2) the difference between the two parties has narrowed over the years as the Dems moved right. On 2), Stonecash shows that the difference between the parties has *widened* over the decades, and that's mostly a function of the Reps having moved very far to the right.
Doug
<http://www.princeton.edu/~bartels/kansasqjps06.pdf>
According to Frank (2004, 242–243), the Democratic Party's "more-or- less official response to its waning fortunes" has been "to forget blue-collar voters and concentrate instead on recruiting affluent, white-collar professionals who are liberal on social issues." This "criminally stupid strategy . . . has dominated Democratic thinking off and on ever since the 'New Politics' days of the early seventies." Meanwhile, the party "has either largely accepted the conservative economic agenda or is perceived to have largely accepted the conservative economic agenda. Either way, economic issues are effectively removed from the table, and social issues are highlighted" (Frank 2005, 12).
I can find no empirical support for the notion that "economic issues are effectively removed from the table" of contemporary American electoral politics. On the contrary, the evidence presented in Table 4 suggests that economic issues have lost none of their potency over the past 20 years, and that they continue to structure presidential voting behavior more powerfully than social issues do. But how can that be the case, if the Democratic Party "has either largely accepted the conservative economic agenda or is perceived to have largely accepted the conservative economic agenda"?
Well, perhaps Frank is wrong about that as well. As it happens, respondents in the 2004 NES survey were asked to place the Democratic and Republican parties on several of the same issue scales that they used to report their own policy views, including both economic issues (government spending and services, government jobs, government aid to blacks) and social issues (abortion, gender roles). Those comparisons are presented in Table 6. The first column of Table 6 reports the average position of white voters without college degrees on each zero- to-one conservative-to-liberal issue scale. The second and third columns report the average position assigned by those voters to the Democratic and Republican parties, respectively. The fourth column measures the average distance between the two parties. The fifth column shows the Republican Party's advantage or disadvantage with respect to average proximity on each issue, while the final column shows the plurality of voters who saw themselves as closer to the Republican Party than to the Democratic Party.8
Looking at what Frank's white working-class voters actually say about where the two parties stand on major issues reveals at least two significant additional problems for his account of the contemporary partisan landscape.
First, his white working-class voters certainly do not see the Democratic Party as having "largely accepted the conservative economic agenda" (2005, 12). Their average placements of the party are well to the left of center on every economic issue – if anything, a bit further to the left of center than their placements of the Republican Party are to the right of center. Nor do they see "[n]o choice" between the parties on economic issues (Frank 2005, 14). On the issue of government spending and services, the average gap between voters' placements of the two parties amounts to almost 40% of the total length of the scale. The same average gap appears on the issue of government jobs; Frank's working-class white voters clearly discern a very substantial difference between the parties' positions on the issue, though his account requires seeing "politicians of both parties in rough agreement" (Frank 2005, 13).
Second, Frank's white working-class voters were neither liberal in absolute terms nor closer to the Democratic Party than to the Republican Party on economic issues. On the central issue of government spending and services, voters who saw themselves as closer to the Republican Party outnumbered those who saw themselves as closer to the Democratic Party by four percentage points. On the issues of government jobs and aid to blacks the pluralities seeing themselves as closer to the Republican Party were even larger – nine and 15 percentage points, respectively. Moreover, 60% to 85% of the voters who perceived differences between their own position and the Democratic Party's position on each of these economic issues said the Democratic Party was too liberal, not too conservative. Thus, it is hard to see why taking even more liberal positions on these issues, or stressing them more heavily, would help the Democrats win back the white working class.
On the other hand, the only two issues on which Frank's white working- class voters did see themselves as closer to the Democratic Party than to the Republican Party were the two social issues in the table, abortion and gender roles. In the case of abortion the advantage is quite modest and not statistically reliable; nevertheless, voters who saw themselves as closer to the Democratic Party outnumbered those who saw themselves as closer to the Republican Party by nine percentage points. For the item on women the Democratic advantage is more substantial, and Frank's white working-class voters actually saw themselves as more liberal than either party.
On the most potent cultural issue in Table 5, gun control, the 2004 NES survey did not include placements of the parties but did include placements of the presidential candidates. On that issue, too, white voters without college degrees saw themselves as more liberal than either presidential candidate. (Four percent said it should be easier to buy a gun; 48% said it should be more difficult.) The big effect in Table 5 implies that these people were more likely, other things being equal, to vote Democratic, not Republican.
The pattern here is exactly the opposite of the one suggested by Frank's account. On economic issues, where the Democratic Party is supposed to have squandered its natural advantage among working-class voters because it has "largely accepted the conservative economic agenda" (Frank 2005, 12), white voters without college degrees see themselves as closer to the Republicans precisely because the Democratic Party is too liberal. But on cultural issues, where Democrats are supposed to "have left themselves vulnerable" to Republican inroads by catering to liberal white-collar professionals (Frank 2004, 245), Frank's white working-class voters say that they, too, are closer to the Democrats than to the Republicans – and sometimes more liberal than either party!