Disclaimer. Don't rely on these old notes in lieu of reading the literature, but they can jog your memory. As a grad student long ago, my peers and I collaborated to write and exchange summaries of political science research. I posted them to a wiki-style website. "Wikisum" is now dead but archived here. I cannot vouch for these notes' accuracy, nor can I even say who wrote them. If you have more recent summaries to add to this collection, send them my way I guess. Sorry for the ads; they cover the costs of keeping this online.
Carsey and Wright. 1998. State and national factors in gubernatorial and senatorial elections. AJPS 42 (July).
Carsey and Wright generally agree with Atkeson and Partin's (1995) "functional responsibility" hypothesis, but they challenge them on a few methodological points; a replication and reanalysis leads to modified conclusions.
Carsey and Wright (1998a) challenge Atkeson and Partin (1995) on several points, prompting a reply from Atkeson and Partin (1998) and a rejoinder from Carsey and Wright (1998b).
Atkeson and Partin (1995) used ANES data to do a state-level analysis. The problem is, the ANES uses a small number of primary sampling units (PSUs) in each state. Although these PSUs produce a good national sample, they do not produce good state-level samples. Atkeson and Partin ought to have done two things to control (partially) for the clustered and pooled nature of the data. First, they should have included state dummies; second, they should have used robust standard errors (since the nature of PSUs raises the probability of correlated residuals within each state).
Second, ANES data suffers from a misreporting problem. Unlike exit polls, ANES respondents don't respond to questions until a few weeks have passed since the election. As a result, people are more likely to recall voting for the winner than is the case. This misreporting bias weakens the apparent effect of presidential approval on voting. Carsey and Wright crossvalidate the ANES findings with similar tests using exit poll data and come to different conclusions.
Carsey and Wright first replicate each analysis, then add state dummies, then run the same tests on exit poll data.
'Senatorial elections': Atkeson and Partin found that only presidential approval matters. Turns out that the strength of the national economy matters too.
'Gubernatorial elections': Atkeson and Partin found that only state economics matter. Turns out that presidential approval matters too.
Like many analysts, Carsey and Wright expect to find that presidential approval affects gubernatorial and senatorial elections; "For many citizens, political judgments are general indictments or rewards of the party in power, usually defined as the presidential party" (995). What about strategic entry? See Brown (2007).
Research by the same authors
Research on similar subjects