In Australia it's well-known (see, for example, here) that a party that holds power at Federal level will - all other things being equal - tend to hold a disadvantage in State elections.

Similarly, it is often observed that the party of the President tends to suffer in American mid-term State elections, for example Gubernatorial elections.

Is there a term for the trend?

  • 3
    The only term I can think of is “protest vote”, which is mainly used in Australia for byelections of seats held by the incumbent party: en.m.wikipedia.org/wiki/Protest_vote
    – Golden Cuy
    Nov 5, 2018 at 9:41
  • In the US, the party of the President is expected to suffer in both federal and local elections. This is the result of the combination of factors: two-party system, incumbency disadvantage, nationalization of local politics, etc. Nov 5, 2018 at 11:32
  • The word that comes to mind is "linkage", the fate of down ticket members of a political party is linked to that of the people at the top of the ticket. In many European countries linkage is even tighter with strong connections between even local government officials and the national political party, and votes for village council seen as an expression of partisan support without much regard to who is actually running in the same was as in an election for the national parliament.
    – ohwilleke
    Nov 10, 2018 at 16:44

2 Answers 2


Short answer

There is no term that describes a trend where a political party does worse in State elections when holding power Federally, because there isn't such a trend in the case of "all other things being equal".

Long answer

I have to disagree with the premise of your question (at least from the Australian perspective). That is, I disagree that the party that holds power at Federal level will - all other things being equal - tend to hold a disadvantage in State elections. [Bold emphasis mine]

Typically, the party that holds power federally will win some of the State or Territory elections that are held while it's in power federally, and it'll also lose some. The main differentiation gets down to local (i.e. State) politics/issues.

There may be a small disadvantage (particularly in the case of a very unpopular federal government), but this doesn't apply across the board in all States and Territories and is much more present in the case of by-elections by virtue of protest votes, although voters may use a State/Territory election to send a message to the Federal government if it's unpopular.

I've read the article you cite (The Cycles of Party Politics, April 2012) but don't see how this supports your assertion. Perhaps I've missed something?

In terms of the United States, the last 25 mid-term elections (including this week) show:

  • There has been a split result in four mid-term elections (1930, 1982, 2010, 2018)
  • The party holding the presidency won both houses of congress in 10 mid-term elections (1922, 1926, 1934, 1938, 1942, 1950, 1962, 1966, 1978, 2002)
  • the party holding the presidency lost both houses of congress in 11 mid-term elections (1946, 1954, 1958, 1970, 1974, 1986, 1990, 1994, 1998, 2006, 2014)

So, in terms of actual seats in congress, there is no real trend - except perhaps the fact that there may be a trend in terms of a split result in Congress now being more likely (twice in past decade, but only four times in the past 100 years).

While I haven't looked at Gubernatorial elections specifically1, looking at the Senate results for the past 25 mid-terms shows us that the party holding the presidency has won the 'States House' (i.e. the Senate) in 12 mid-term elections (and lost it in 13), so this also doesn't denote a trend.

1. The issue of partisan Gerrymandering in a number of states means any such analysis of Gubernatorial elections is fraught with danger unless allowances are made on a state-by-state basis for the impact of any such gerrymandering.

  • Searching a bit more on this topic I came across this post by one of Australia's better-known psephologists, who claims to provide evidence that what he calls "Federal Government drag" exists. He cites the graph in the original link I posted as evidence for such a drag. Nov 10, 2018 at 8:05
  • 1
    Thanks for the reference. When I get a chance (maybe tomorrow) I’ll take a look. Might also see if Malcolm Mackerras has written anything on the subject. In the meantime, I’ve made some minor edits to my answer to clarify it a little.
    – Monomeeth
    Nov 10, 2018 at 9:27

In the process of responding to another post in this thread, I came across a potential answer from the Australian psephologist Kevin Bonham, who writes here (backup here) about federal government drag.

From the context it seems apparent that Kevin was coining this term, and so I'll wait and see if someone can show that there is another name for this concept that pre-dates 2014.

There is some very interesting content in this thread suggesting that the concept of Federal government drag may be flawed. That is hard for me to assess, but it seems that at least some psephologists (not just the post I linked, but some I've talked to in person) believe federal government drag really exists, at least in an Australian context. Hence it still seems reasonable for the concept to have a name.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .