Research
Articles
WILLIAM D. BERRY,
Florida State University
EVAN J. RINGQUIST,
Indiana University
RICHARD C. FORDING,
University of Kentucky
RUSSEL L. HANSON,
Indiana University
A Rejoinder: The Measurement and Stability of State Citizen Ideology
Our
initial article challenged three of Brace et al.'s (2004) claims: (1)
state citizen ideology is stable over time; (2) the Berry et al. (1998)
measure of state citizen ideology is invalid; and (3) assessing the effect
of state citizen ideology on state policy requires analysts to first parse
out national or across-the-board changes in ideology. By stating that
"the authors of the preceding article reinforce most of our substantive
points," Brace et al. (2007, 000) leave the impression that we agree with
these claims. We do not. In fact, we see no evidence in this exchange
to support any of these claims or to rebut our criticisms of them. To
substantiate their first claim Brace et al. (2004) report that residuals
from fixed-effects regression models of two measures of state ideological
self-identification display few significant time trends. In this issue,
we show that the results of the residual analysis are almost certainly
artifacts of low statistical power and high measurement error and, hence,
cannot support their claim. Brace et al. Defend their second claim by
showing that Berry et al.'s measure of citizen ideology both produces
far more statistically significant time trends than do the two measures
of ideological self-identification and is weakly correlated with these
two measures. In response, we once again point to the artifactual nature
of their regression residual results and note that Brace et al. (2001)
report incorrect correlations. These empirical criticisms are independent
of the distinction we draw between policy mood and self-identification;
were we to abandon this distinction completely, which we recommend against
on theoretical grounds, these criticisms of Brace et al. (2004) would
still apply. We reject Brace et al.'s third claim—that changes in
state citizen ideology associated with national trends cannot meaningfully
influence state public policy—on logical grounds, noting that across-the-board
changes in state ideology may very well produce changes in state public
policy.
|
|
