It’s all gated, but Critical Review is worth a subscription. I found Zaller’s own entry the most interesting. Some excerpts (each paragraph is plucked at random–they are not a sequence):

An ideology is a set [of] policy positions recommended by informal coalitions of political pundits, intellectuals, and interest-group representatives…The purpose of ideology is to persuade citizens at large, and espectially the more politically active segment of the populace, of what ought to be done in politics…different people are attracted to liberalism and conservatism for different reasons…what Conover and Feldman call symbolic attachments–e.g. disliking “Big Business,” liking “Women’s Liberation”–are more closely associated with evaluations of liberalism and conservatism than are policy preferences.

If there is one thing that my “political education” over the last 20 years has taught me, it is that one cannot tell a sensible story about public opinion and democracy in the United States without ascribing a central role to interest group and activist policy demanders.

Parties offer policies that are acceptable to their policy-demanding activists and calculate to appeal to particular voting blocs. There is no expectation that parties…will offer policies simply because voters want them. Nor does the median voter’s position…play a significant role. Majorities obtained through any means consistent with the agendas of policy demanders are what parties care about.

According to [Larry M.] Bartels’ analysis, each term in office (after the first) [for one political party] offsets 1.29 percentage points of Q14/Q15 growth [in real disposable income in the spring and summer of the Presidential election year].

The Bartels model raised a number of interesting questions to me. As I read the chart in Zaller’s paper, a first-term-for-the-party incumbent is likely to be re-elected as long as real disposable income does not fall the spring and summmer of the election year. That seems like a low bar. The finding that the bar gets higher the longer the incumbent party has been in office could be due to a combination of two things. One is that the voters get tired of the incumbent party, regardless of what the other party does. The other possibility is that the longer a party is out of power, the more desperate it comes, and the more willing it is to adapt in order to win. I think of those two possibilities as having rather different implications.

The vast amount of government policy-making…that is beneath the radar for most voters…is ceded to the demands of interest groups and activists…Where does the sending of cues by partisan leaders fit into this model of respresentation? It doesn’t have a primary role. Its secondary role is to increase political harmony within coalitions by gaining the assent of members for the common agenda.

My three-axis model would describe part of the cueing process. A politician who wants to send a “cue” to progressives can talk about issues along the oppressor-oppressed axis. A politician who wants to send a “cue” to conservatives could talk in terms of civilization vs. barbarism. A politician who wants to send a “cue” to libertarians would talk in terms of freedom vs. coercion.

Note that Republicans have a more complicated problem, because they want to send cues to conservatives and libertarians.

Note that all of the exercises in mobilizing voters, whether using group-identity or three-axis cues, or other means, are simply for the purpose of winning elections. Once in power, politicians primarily serve interest groups. That is why an ideologically committed voter always feels keen disappointment with how little is accomplished when his or her preferred politicians win.

In 1971, President Richard Nixon [imposed] emergency wage and price controls…[in] a survey of Republican activists…Support for price controls was 37 percent before the speech but 82 percent afterwards

My guess is that a survey of Democratic activists on NSA snooping would show a similar before-and-after. That is, before it was revealed that the NSA was snooping on Americans on Obama’s watch, a small percentage of Democratic activists would have favored such snooping. Now, I conjecture, a much larger percentage of Democratic activists favors such snooping.