Abramowitz writes:

Krugman claims that the relationship between frequency of church attendance and Republican voting is much weaker among lower income white voters than among upper income white voters. This claim is not supported by evidence from the 2004 national exit poll. According to the exit poll data, the relationship was equally strong among lower and upper income white voters. Among white voters with family incomes of less than $30 thousand, George Bush was supported by 68% of those who reported attending church more than once per week vs. 33% of those who reported never attending church.

(Emphasis supplied.) Is that what Krugman claimed? No. Here is what Krugman wrote:

It’s true that people in poor states are more likely to attend church regularly than residents of rich states. This might seem to indicate that faith is indeed a response to economic adversity.

But this result largely reflects the fact that southern states are both church-going and poor; some poor states outside the South, like Maine and Montana, are actually less religious than Connecticut. Furthermore, within poor states, people with low incomes are actually less likely to attend church than those with high incomes. (The correlation runs the opposite way in rich states.)

What is demonstrably false is that Abramowitz accurately characterized Krugman's argument. Krugman said that IN POOR STATES (now quoting Abramowitz) "the relationship between frequency of church attendance and Republican voting is much weaker among lower income white voters than among upper income white voters." Abramowitz's rectiing 2004 NATIONAL EXIT POLLS is a complete nonsequitor to Krugmans's factual assertion. It is either disingenuous or dishonest of Abramowitz to pretend Krugman wrote something he did not. Thus, Abramowits point one refutation is utter nonsense.

Abramowitz then writes:

Citing the research of Princeton political scientist Larry Bartels, Krugman claims that Democrats do better among working class voters now than they did during the 1960s and that the success of the Republican Party in American politics since that time is entirely explained by GOP gains in the South. But this is simply not the case. Republican gains in the South are only part of the story of GOP success since the 1960s.

At this point, we know we can not trust Abramowitz's characterization of Krugman's argument so we must go to the source material to see what Krugman actually argued. Krugman wrote:

In his Op-Ed, Mr. Bartels cited data showing that small-town, working-class Americans are actually less likely than affluent metropolitan residents to vote on the basis of religion and social values. Nor have working-class voters trended Republican over time; on the contrary, Democrats do better with these voters now than they did in the 1960s.

It’s true that Americans who attend church regularly are more likely to vote Republican. But contrary to the stereotype, this relationship is weak at low incomes but strong among high-income voters. That is, to the extent that religion helps the G.O.P., it’s not by convincing the working class to vote against its own interests, but by producing supermajorities among the evangelical affluent.

(Emphasis supplied.) Abramowitz does not even touch Krugman's actual argument. Instead he engages in obfuscation:

Democratic identification (including leaning independents) among white voters outside of the South fell from 50% to 44% between 1962-70 and 2002-2004 while Republican identification rose from 45% to 51%. Thus, an eight point Democratic advantage during the 1960s was transformed into a seven point Republican advantage in 2002-2004.

But Republican gains were much larger among two key Democratic constituencies. Among northern white Catholics, Democratic identification fell from 65% during the 1960s to 44% in 2002-2004 while Republican identification rose from 26% to 49%. Thus a 39 point Democratic advantage among northern white Catholics was transformed into a five point Republican advantage.

Do you see the misstatement Abramowitz made? He treat ALL white voters as poor and all Catholic voters as poor. We know this is wrong. Catholics in the United States have been upwardly mobile economically. Nothing Abramowitz cites in any way refutes Krugman and Bartels' factual points. Again, what is demonstrably false is that Abramowitz addressed Krugman's factual points.

Abramowitz continues:

Similarly, among northern white blue collar workers, Democratic identification fell from 61 percent during the 1960s to 41 percent in 2002-2004 while Republican identification rose from 31 percent to 48 percent. Thus, a 20 point Democratic advantage among northern white blue collar workers was transformed into a seven point Republican advantage.

This is the closest Abramowitz comes to addressing Krugman's point. But two things stand out - first, Abramowitz provides no economic definition of what "blue collar worker" means. We do not know what incomes he is talking about. Second, Abramowitz leans on "party identification" rather than voting patterns. This simply is not enough to declare that Krugman's factual assertions are "demonstrably false."

Indeed, it is instructive to consider what Krugman actually wrote:

So why have Republicans won so many elections? In his book, “Unequal Democracy,” Mr. Bartels shows that “the shift of the Solid South from Democratic to Republican control in the wake of the civil rights movement” explains all — literally all — of the Republican success story.

Abramowitz's response is to distort:

It is clear from these data that the Democratic Party's problems in recent years have not been confined to white voters in the South. Democrats have also lost ground among white ethnic and working class voters outside of the South, voters who were once crucial components of the party's electoral base. While Barack Obama's formulation of the problems that the Democratic Party has been having with these voters may not have been artful, there is no doubt that the problem is real.

This was not Krugman's point. Krugman said that the reason why Republicans WON was confined almost entirely in the South. And anyone familiar with a Blue State/Red State map knows this to be true. The South flipped to the Republicans in the wake of the civil rights movement. Tom Schaller and others, including me, have argued for what I have termed a Lincoln 1860 strategy in the short term, win without the South. Schaller rightly points out that Obama has real strengths in the West and the Midwest that make him a better candidate for executing that strategy. But Abramowitz was intent on misstating Krugman's factual assertions and his argument. He missed the obvious opportunity to tout Obama's candidacy. Why? Because disagreeing with St. Barack Obama is an unpardonable offense.