Five Lessons from the Metro Mayor Elections

Five Lessons from the Metro Mayor Elections

Back in April, I tried to consider what our expectations of the 8 combined authority (metro) mayoral elections should be. As I said at the time, these weren’t meant to be predictions, but an attempt to think about what a “baseline” result would look like, based solely on national polls and general election results. On the whole, the elections weren’t that unexpected, but I wanted to return to my blog to see what I got wrong.

Overall, I thought Labour should go into the elections with the expectation of making gains. In the end, this was broadly right. Labour gained two mayoralties from the Conservatives – West of England and Cambridgeshire and Peterborough – while also adding the newly created West Yorkshire. Given those wins, the Metro Mayor races were some of the most positive results of Labour’s poor showing on 6th May.

Lesson 1: Incumbency is BIG but inconsistent

However, all of those Labour gains came in elections without an incumbent. I identified measuring the incumbency effect as one of the most important reasons to consider our “expected” results before the election and this was vindicated by how bad my guesstimate of the incumbency effect was.

I wrongly assumed that the incumbency effect would be bigger in London than other regions, suggesting a bonus of somewhere between 1.75 and 6.3%. Instead, it seems that incumbents performed better outside London. In the 9 Metro Mayor elections since 2000, incumbents have performed an average of 9.8% better than would be expected using just polls and general election results. Excluding London, incumbents did an average of 11.2% better.

This effect is much larger than I expected but it is also inconsistent. Some incumbents massively overperformed. Ben Houchen, for example, performed 20.7% better than in Hartlepool than the Conservative candidate for the Hartlepool by-election, held on the same day. Meanwhile, other incumbents struggled. James Palmer did about 9% worse than expected in Cambridgeshire and Peterborough, Steve Rotherham did about 3% worse in Liverpool, and Sadiq Khan did around 2% worse in London. Unlike Hartlepool, these numbers are comparing against General Election results rather than elections held on the same day, but they still indicate that incumbency doesn’t equate to success.

James PalmerConCambridgeshire and Peterborough (2021)-9.0%
Steve RotherhamLabLiverpool (2021)-2.9%
Sadiq KhanLabLondon (2021)-1.9%
Ken LivingstoneLabLondon (2004)+2.7%
Ken LivingstoneLabLondon (2008)+10.9%
Andy StreetConWest Midlands (2021)+15.8%
Boris JohnsonConLondon (2012)+20.1%
Andy BurnhamLabGreater Manchester (2021)+23.7%
Ben HouchenConTees Valley (2021)+28.4%
Incumbent mayor performance compared to expectations based on polls and general election results

As a lot of people pointed out before the election, Houchen’s huge win seems to have come off the back of unprecedented support from central government, for example the nationalisation of Tees Valley Airport and the announcement of “Freeport” status. Incumbency for Houchen in a key Tory electoral battleground probably means something quite different to a Labour mayor, or even a Tory in a less electorally potent region.

Lesson 2: Labour struggles in the first round

Another lesson from 2021 is that Labour performs worse than the Conservatives or Liberal Democrats on first preferences than you’d expect. On average since 2000, with the effect of incumbency removed, Labour has performed 2.8% worse than we would expect based on General Election results and polls. The Conservatives performed 1.7% worse, while the Liberal Democrats did 3% better.

This is likely caused by votes for third parties. It is well established that smaller parties perform better in ‘second order’ and preferential elections, and this has been the case in Metro Mayor elections. It is likely that this disproportionately affects the Labour vote, especially since the collapse of UKIP.

Anecdotal reports also suggest Labour campaigns in target Mayoralties were underfunded and chaotic, with lower quality candidates than hoped.

Lesson 3: Conservatives struggle in the second round

Meanwhile, the Conservatives performed very poorly in the second round. On average, in 2021, Labour won 62.8% of second preferences in the five elections which went to a second round, while the Conservatives got an average of 37.2%. This compares to an average of 45.9% for the Conservatives in Metro Mayor elections before 2021, possibly reflecting the isolation of the Conservative Party from other parties as it has absorbed the 2016 Leave vote.

Cambridgeshire and Peterborough30.7%
West of England31.3%
West Yorkshire39.9%
West Midlands45.7%
Percentage of second preferences received by Conservative candidates 2021

Most elections using Supplementary Vote are not close enough that the second round changes the result but going forward we should expect more mayoral elections like that in Cambridgeshire and Peterborough where the Conservatives were defeated in the second round after winning the first. Right on cue, the Conservative are planning to abolish SV altogether…

The second round advantage for Labour almost certainly outweighs the electoral system’s possible cost on the party’s first preference vote.

Lesson 4: ‘Fourth’ parties performed worse in 2021

Ever since Ken Livingstone’s successful independent bid for London Mayor in 2000, parties aside from Labour and the Conservatives have struggled to make a dent in Metro Mayor elections. Even the Liberal Democrats have only made it to the second round once in 16 elections.

Although the Greens did impressively well in the West of England, winning 21.7% of the vote and coming within 7% of the second round, parties aside from Labour, the Conservatives and the Liberal Democrats did comparatively worse in 2021. In elections where one more more ‘fourth party’ candidate stood, they got an average of 15.8% combined, compared with 19.4% for previous elections.

This might be due a disproportionate effect of COVID on candidates without established party backing, or the fact that a higher proportion of 2021’s elections were outside London, which seems to attract a higher number of minor candidates.

Lesson 5: These elections can be weird

Finally, an important lesson is that these elections throw up surprising results. While overall, Labour performed poorly in first round elections and incumbency provided a big boost – Cambridgeshire and Peterborough seemed to experience the opposite. In Tees Valley, Ben Houchen won an unprecedented landslide in an area which Labour had only narrowly lost at the 2019 election. In London, where journalists had spent most of the campaign mocking Shaun Bailey’s Conservative campaign, the result was far closer than expected.

In terms of trying to model future elections, this means dialling up the uncertainty. We should expect Metro Mayor elections to be relatively much more competitive than Westminster elections in those regions.

This is undoubtedly good for local democracy, which is usually the victim of one-party fiefdoms and practically uncontested elections. It also provides an opportunity for very different styles of politician to make a name for themselves outside of Westminster, something which England has been lacking compared with other UK nations.

What’s next?

So, given these lessons, how should we consider the next set of Metro Mayor elections? There are only two combined authorities electing Metro Mayors between now and 2024. These are South Yorkshire, currently held by Dan Jarvis MP, in 2022, and North of Tyne, currently held by Jamie Driscoll, in 2023.

If Jarvis and Driscoll stand again, it will be interesting to see if the current trend of a smaller incumbency effect for Labour mayors continues. Both areas were won by Labour in 2019, South Yorkshire by 10% and North of Tyne by 5.6%, so Labour should probably be fairly confident of holding them.

If they were held today, with Jarvis and Driscoll running for reelection, Labour would be the heavy favourites (96% and 93% probability, respectively). With no incumbent, the races would be much more competitive, with Labour having a 82% chance in South Yorkshire and 73% in North of Tyne.

That said, anything could happen by next year, let alone 2023, so maybe I will look back and laugh at this when the Yorkshire Party surges to a crushing victory.

(Featured image by Ввласенко)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create a website or blog at

%d bloggers like this: