Despite the best efforts of a tax forum and media tail-chasing about the Labor leadership, this has been an unusually sedate week in federal politics.

This presumably can be put down to a national hangover after a long weekend that brought grand finals in both major football codes, together with a related consequence: Newspoll did not take the field over the weekend as per its usual schedule, thereby sparing the government its regular fortnightly morale blow. However, the polling gods found an alternative means to maintain the water torture Labor has suffered since early this year, courtesy of a Newspoll of state voting intention from Western Australia.

The kicker here was yet another primary vote figure with a two in front of it — 29%, to match the 26%, 28% and 27% Labor recorded in the most recent federal, Victorian and Queensland polls, and the 25.6% it was able to scratch together at the NSW election. The South Australian government was only able to escape the same fate by making it to 30%. Labor governments in Victoria and South Australia were performing roughly as badly following bank collapses at the turn of the 1990s, but the results federally and in NSW, Queensland and Western Australia are without precedent for Labor since Newspoll first opened for business nearly three decades ago.

The uniformity of the results has inspired talk of permanent damage to a “Labor brand”, a notion that presumes a common source to electoral effects at federal and state level. To consider the extent to which this is so we can turn to the archives of Newspoll, which has been publishing regular polling figures federally and for each mainland state since 1985. The chart below is based on quarterly voting intention in federal and state polling, the aggregated state result giving weight to each state’s figures in proportion to its population.

What emerges is a remarkable tendency for the two lines to move in synchronicity over the short term. However, a longer-term view reveals a series of distinct phases in the relationship between federal and state voting intention. For the first few years after 1986 there is a consistent gap in favour of federal Labor of about 4%. This vanishes in late 1990, when the federal Coalition under John Hewson begins to record consistently strong poll leads for the first time since the election of the Hawke government. This shift brings the two lines almost perfectly into line, where they will broadly remain until late 1999.

It is at this point that Labor’s state-level ascendancy in the 2000s becomes clearly established: Victorian and Queensland governments are in the first flush of their honeymoons, while the Carr government establishes a harder fought ascendancy in New South Wales. Labor thereafter remains stronger at state level until the federal party catches up on Kevin Rudd’s watch after December 2006.

The election of the Rudd government in November 2007 can be seen as initiating a new episode of alignment in federal and state fortunes, albeit one separated by a two-year gap. At state level, Labor’s fortunes immediately enter the decline that leads to its present all-time low; federally, the party at first scales new heights under Rudd before entering a precipitous descent at the start of 2010. The state figures meanwhile continue plunging at a similar clip, maintaining a constant gap between the two of roughly 5%.

One reading of this is that federal Labor began chasing state Labor’s tail as soon as the Rudd government’s honeymoon was over, and certainly the conventional wisdom would have it that exactly this happened in New South Wales. However, this seems to turn on its head the experience of the Howard years, when the consensus was that the Coalition’s dominance at federal level was damaging it in the states.

Either diagnosis seems unduly deterministic in any case. Labor’s current tail-spin is clearly better explained in terms of mutually reinforcing effects at federal and state level, caused in different ways by self-inflicted political difficulties and an unprecedented unalignment of federal and state electoral cycles.