For a couple of weeks there, those of us who follow opinion polls a little more closely than we should felt that the inflation in Labor support after the May 13 budget was losing a bit of air.

That’s looking less clear now, after four new polls this week which covered the full methodological gamut:

  • Labor’s lead was back to its 55-45 peak in the latest Newspoll, after falling to 53-47 the previous fortnight. The poll was conducted for The Australian from Friday to Sunday, by live interviewers contacting 1161 respondents by landline phone.
  • Reflecting a long-established pattern, the fortnightly Roy Morgan poll leaned still further to Labor, showing an eye-catching two-party preferred result of 57.5-42.5. However, this comes down to 54.5-45.5 on the more conventional two-party preferred measure, which allocates minor parties’ preferences as they flowed at the previous election. This poll was conducted by face-to-face and SMS over the last two weekends from a combined sample of 2797.
  • Labor’s lead in this week’s Essential Research poll was steady at 52-48 — a stronger result than it looks given that Essential publishes a fortnightly rolling average, and the previous week’s result had Labor down two points. This suggests that Labor’s weakness in last week’s sample wasn’t repeated this week. Essential’s weekly samples encompass around 1000 respondents from a volunteer panel of 100,000, who are surveyed online.
  • A ReachTEL poll for the Seven Network had Labor’s lead at 53-47, compared with 54-46 at its last poll two weeks before the budget. This was an automated phone poll of 3376 respondents, conducted last Thursday.

A trend measure of all published federal polling, known to readers of my blog The Poll Bludger as BludgerTrack, records Labor increasing half a point on last week’s reading to reach a two-party lead of 53-47, while an accompanying seat projection has them one shy of Kevin Rudd’s total of 83 seats at the 2007 election.

Observed over time, the blunting of the recent turn to the Coalition means the post-election trend can again be understood in three clear phases.

The first covers the Coalition’s unusual failure to enjoy a post-victory honeymoon, with the polling conducted in September and October broadly in line with the result at the election.

The onset of the second phase roughly coincides with the government’s Gonski debacle in late November, which was perhaps reflected in this week’s finding from Essential Research that its architect, Education Minister Christopher Pyne, was the least popular of the government’s seven most senior ministers.

Labor thereafter had its nose in front until the second landslip after the budget, when its two-party vote geared up to the low to mid-50s.

For all that may have been claimed for the impact of issues prevailing in the news cycle at any given time, shifts that have been noted in the polls other than these two have probably been little more than statistical noise.

Beneath the two-party surface, the other notable trends have been the steady ascent of the Greens, up from 8.5% at the election to around 11% now, and more recently the Palmer United Party, which experienced a post-election lull before surging to the 6-7% range after the WA Senate election and the budget.

It appeared in the later part of June that Palmer United may have taken a bit of a knock after Mike Willesee’s unflattering and widely discussed report on the party aired on the Seven Network’s Sunday Night program.

However, this week’s numbers have it back on track, perhaps reflecting the publicity the party has received as its senators prepare to take their seats — although both the earlier dip and the more recent recovery have been inside the margin of error.

Regrettably, Palmer has piked on an opportunity to put his electoral strength to a real-world test on his party’s home turf, having declined to field a candidate at Queensland’s Stafford byelection on July 19.