Through the recording of response times in a national four-wave bilingual p
anel survey, this study reports improvements in the prediction of vote choi
ce rep to I year in advance of a federal election. These results were achie
ved with conventional computer-assisted telephone interviewing (CATI) softw
are, indicating that the immediate Else of response time measures is both p
ractical and attractive for commercial as well as academic survey units. Ev
en so, response latencies were found to be sensitive to political circumsta
nce, such that timings should be analyzed separately for minority and major
ity populations. Moreover, a broad analytic focus, beyond timing only vote
intention and partisan commitment, is recommended because latency data on c
ore questions of identity and allegiance reveal a great deal about the cont
ours of political context.