A short honeymoon for Starmer?
Voters' love for new governments doesn't last
Yesterday was an unwanted milestone for the new Labour government, as BMG published the first poll to show them trailing the Conservative opposition, albeit by a narrow 29-28 margin in a fragmented field. Labour led the Tories in opposition for two and a half years, their lead in government has lasted less than four months. Some honeymoon.
Labour advisors who have grown used to commanding poll leads in opposition can take some comfort from history. The swift Starmer slump is not unusual. Voters seldom give new governments much credit for long. All but three of Britain’s nineteen post-war governments have trailed in the polls within a year, and most (fourteen of nineteen) have fallen behind within six months.
Early rejection by the electorate is a rite of passage for new administrations. Only two buck the trend. Clement Attlee’s post-war Labour administration was favoured for just over two years, while Tony Blair is the true outlier on this poll measure, as on so many others, with unbroken poll leads for well over three years (and even his first deficit was a brief blip driven by fuel protests in 2000).
Days until first poll deficit, governments 1945-2024
Source: Mark Pack PollBase
The 118 days to Starmer’s first poll deficit is below the 193 day average for all governments, but that average is inflated by the two outliers and by the lower frequency of polling in the early post-war decades. Mark Pack’s pollbase dataset records an average of one or two polls a month in the 1950s and 1960s. Sparse polling likely delayed the first public rejections of Attlee and other early post-war prime ministers, pulling up the average honeymoon period.
While falling behind is never great for morale, Labour aides can take some comfort from knowing that most of their predecessors also trailed early on, and many went on to win again. Macmillan, Wilson, Thatcher, Major and Cameron all recovered from early polling setbacks to win subsequent elections.
Governments don’t get long honeymoons, but do Prime Ministers do any better? Here we have less history to draw on, as the longest continuous measure of leader approvals, run by IPSOS-MORI, only goes back to 1977. So we can only look at how voters have reacted to new Prime Ministers from Margaret Thatcher onwards, shown on the chart below.
Days until first deficit in leader approvals, governments 1977-2024
Source: IPSOS-MORI satisfaction ratings archive (Starmer’s final figure is not yet known)
New Prime Ministers don’t get a long honeymoon either, with most trailing their opponent in MORI’s approval ratings within a year of taking office. Tony Blair is an even more freakish outlier on this measure - voters gave him stronger ratings than his various Tory opponents for a massive six years, more than triple the figure for his nearest rival (and predecessor) John Major. Blair’s 2243 days of approval leads is a whopping ten times the average figure for all the other Prime Ministers since 1979 (218 days). Tony Blair is truly the Don Bradman of Prime Ministerial polling.
What brings Prime Ministerial honeymoons to a close? Sunak took on a poisoned chalice, so fell behind almost immediately.1 Truss was the victim of the economic turmoil unleashed by her “mini-budget”, which saw her rating crash as soon as polling capturing voter reactions was published. Theresa May’s ratings collapsed during her disastrous 2017 election campaign, which saw her fall behind Jeremy Corbyn for the first time. Thatcher’s honeymoon, unusually, was ended by continued voter affections for her defeated opponent James Callaghan, who remained in office as leader of the opposition.
It is oppositions, however, who end most Prime Ministerial honeymoons, by installing a new leader who voters decide (at least at first) they prefer. The elections of Keir Starmer, Ed Miliband, David Cameron and John Smith all ended Prime Ministerial honeymoons. Given Starmer’s already mediocre approval ratings, Kemi Badenoch is more likely than not to be the fifth name on that list, but as Ed Miliband can tell her, early advantages need not last particularly for leaders weighed down by an unpopular, newly rejected party. The window of opportunity with voters may yet close swiftly for the new opposition leader, as it has for the new Prime Minister.
This piece was edited to address an error in the first version, which stated Margaret Thatcher led in the polls for two years after taking office in 1979. In fact, she fell behind in her very first polls, less than two months in. My thanks to Mark Pack, creator of the PollBase database, for swiftly spotting and alerting me to this error.
This is a rare polling measure where Liz Truss doesn’t finish dead last. This is largely a matter of luck - MORI’s political monitor for September 2022 was in the field from 7-15 September 2022, during the period of mourning for Elizabeth II, who died on 8th September. Even Liz Truss struggled to cause much trouble during a week of mandated political inactivity. She soon made up for lost time thereafter, with the mini-budget which brought things crashing down arriving on 23rd September. In Truss’s second (and final) MORI approval poll, her net rating fell 51 points from -2 to -53.
Two points: a) Starmer has had the shortest honeymoon since George Eliot but fortunately has not ended up in the canal; b) Last month was the 60th anniversary of both Wilson's first win and my joining the Labour Party for the first time. I remember Wilson (in the fashion set by Kennedy) laid great stress on the importance of his first 100 days. In the event those days proved to be a succession of disasters culminating in the loss of the Leyton by-election, a 'safe seat' that was meant to return his Foreign Secretary Patrick Gordon Walker to the Commons after he lost in the infamous Smethwick election in 1964, yet 14 months later he won a landslide victory with a majority of 96. Starmer has got four and a half years yet.
The really interesting thing on the personal popularity chart is the Cameron is the only successor of Blair who had a Premiership longer than Blair’s period (and even then only a fortnight longer)