Two findings from Marc and Jonathan's book strike me as both significant and persuasive. First, authoritarianism is a crucial ingredient in American public opinion on contemporary issues from terrorism to immigration to gay rights. Second, authoritarianism's effects are muted, not exaggerated, by threats, as long as those threats truly threaten all citizens, non-authoritarians and authoritarians alike. In other words, universal threats make non-authoritarians act more like authoritarians, not the reverse.

But I am not ready to go much further than this. Let me push Marc and Jonathan on two points.

First, the book is premised on a historical argument: beginning as early as the late 1950s, Republican Party elites worked to shift the issue agenda toward race, crime, family values, and militarism. Marc and Jonathan suggest that the current potency of authoritarianism derives from these efforts. However, due to data limitations, it is difficult to demonstrate that authoritarianism's role has actually increased. Marc and Jonathan can only show that it mattered in 2004 and 2006. We still do not know whether American public opinion is more structured by authoritarianism now than in the past.

In the one instance where Marc and Jonathan can show change over time, the pattern of change is curious given their narrative. In Chapter 7, they show that authoritarianism is significantly related to partisanship, but only as of 2004. (There is also tentative evidence from 2006). There is no significant relationship in 1992 or 2000. Why, especially if the issue agenda had been shifting toward these new authoritarian issues for 40 years? It's not that these new issues weren't salient in 1992 or 2000: Chapter 5 shows that gay rights were much discussed in 1992 and in 2000, as were "Christian fundamentalists" and similar concepts. I'm left wondering if the association between authoritarianism and party identification is simply an artifact of events between the 2000 and 2004 elections. I took a look at the 2008 survey from the same series that Mark and Jonathan rely on, the American National Election Studies. Among non-blacks in 2008, the correlation between partisanship and authoritarianism is a fairly meager 0.04, where 0 equals no correlation and 1 a perfect positive correlation. By contrast, the correlation between partisanship and ideological self-identification (as liberal, moderate, or conservative) is 0.59. Authoritarianism doesn't seem that strongly associated with partisanship in 2008.

Second, I want to interrogate their definition of polarization and, by extension, the kinds of issues they see as fundamentally linked to authoritarianism. As Paul also noted, here is the definition:

To us, the essence of polarization is when hot-button issues become salient concerns for a large percentage of people. People feel intensely about these issues because they tap into something deep inside them.

Similar language appears throughout the book--e.g., "strong feelings," "visceral quality," etc. I think this characterization of public opinion is too simplified.

Take abortion. People aren't divided into pro-life and pro-choice camps. Most Americans favor abortion under certain circumstances, such as incest or rape. A slim majority of Americans oppose abortion under other circumstances, such as if the woman cannot afford to raise the child. Almost all Americans oppose abortion under still other circumstances, such as for sex selection.

Take gay rights. Americans remain quite divided on gay marriage and adoption. But, as Marc and Jonathan's data shows, they largely support protecting gays from job discrimination and allowing gays to serve in the military. Even two-thirds of the most authoritarian respondents support these ideas.

Does it help us understand opinion on these issues to think of them as "emotional" or "visceral." To me, opinions on these issues don't portend partisan or ideological bloodsport, expect among activists. They suggest fairly nuanced distinctions, not simply "strong responses." And those distinctions exist among authoritarians and non-authoritarians alike.

Because of both reasons, I remain uncertain whether American politics today is somehow truly more "authoritarian." There is no question that authoritarians and non-authoritarians are different, but whether this difference is historically noteworthy remains to be seen. Furthermore, although Marc and Jonathan equate polarization with a "feeling"--e.g., "the differences between Republicans and Democrats have come to feel irreconcilable"--any such "feeling" is difficult to measure, both now and in the past. Quotes from Ann Coulter and Bill Maher certainly seem to make the point, but I'm not sure how much we can generalize from those two.

Marc and Jonathan close with an epilogue from 2008. I will as well. I estimated a simple model of presidential vote choice among non-blacks, including only partisanship and authoritarianism as predictors. The effect of authoritarianism is substantively and statistically significant: respondents who score higher on this scale are less likely to vote for Obama. But what if we reduced authoritarianism? Would that make any difference to the election's outcome? For the sake of illustration, I took the "strong" authoritarians--those who scored above the average on this scale--and made them into "average" authoritarians. I simulated Obama's vote share with and without this change in authoritarianism. Reducing authoritarianism in this hypothetical fashion increased Obama's margin of victory by only a fraction of one percentage point.

Seem surprising? Not really. People who score high on the authoritarianism scale aren't likely to vote for Obama for a host of other reasons.

In other words, if Obama is in trouble in 2012, Democrats won't have spanking to blame.