New data suggests a decline in public confidence in Bristol Council under an elected mayor

In March the Bristol Civic Leadership Project announced that Bristol council’s leadership had become more visible since 2012, presumably because of the Mayoral system. Some politicians celebrated the research as demonstrating the popularity of Marvin Rees’ leadership, while others commented that visibility does not necessarily mean quality. 

Unfortunately the Leadership Project published only a 4 page analysis in 2020, as opposed to 57 pages in 2013 and 65 pages in 2015. Hidden in the announcement was the fact that much of the latest data was not being released. The Leadership Project also chose not to remind the public that they had done three large-scale surveys – in 2012, 2014 and 2018. Instead, they compared only the situation in 2012 and 2018.

The effect was to hide a dramatic decline in public confidence in the city’s leadership between 2014 and 2018.  Since Marvin Rees assumed office, confidence, trust, sense of ability to participate, in fact every single indicator available in the latest results (which can then be compared with the 2012 and 2014 results) shows there has been an extraordinary decline in public perception of the mayor and the city’s leadership under Marvin Rees.

I have taken every piece of the 2018 data which was made available in the 2020 publication, and compared it with the earlier results.  This means, simply – no data cherry-picking. The results are striking.

The most positive aspect of the introduction of the mayor’s role has been to raise the visibility of city leadership.  Yet while there was a big improvement under ex-mayor George Ferguson, things have gone backwards under the current mayor Rees.  Here are the results of polling of a representative sample of Bristol Citizens:

Strikingly, the sense of democratic participation increased under Ferguson, but fewer citizens now feel involved and able to make a difference than before the mayoral system was introduced:

Finally, trust in the council’s decision-making capacity, which improved during the Ferguson era, has dramatically declined to levels inferior even to the supposedly muddled and disunited pre-Mayoral days:

These results reflect polling of ordinary citizens – 658 respondents in 2012, 1013 respondents in 2014 and 680 respondents in 2018.

In addition to consulting ordinary citizens, the Leadership Project consulted three groups of leaders: councillors, community and business leaders, and public management and professionals. 

Councillors have been less than happy with the introduction of the mayoral system, not surprisingly as power has moved away from them towards the mayor.  Still, on a couple of issues there has been a step-decline under Rees, and in no case an improvement.

Strikingly, fewer councillors feel that responsibility for making decisions is clear now than ever before:

Democratic accountability and checks and balances are seen to have dramatically declined under Rees:

More councillors are now unhappy with how well their constituents’ views are represented:

The most gung-ho supporters of the mayoral system were the community and business leaders:

But more of them now feel that the clarity about who makes decisions is lower now than under Ferguson and lower than even in pre-mayoral days:

Particularly marked is the change during Rees’ tenure in the views held by the third group of leaders surveyed, the managers and professionals.  They saw an increase in clarity of leadership under Ferguson, but now see a decline to even pre-mayoral days under Rees. Their views are particularly surprising given that Ferguson operated largely un-assisted for much of his tenure, while Rees has brought in several expensive managers.

The Bristol City Leadership Project did not make their full results available in March, or we could have explored these issues in more depth.  Why they chose not to is unclear – possibly, they wanted their audience to concentrate on contrasting “before the [institution of the] Mayor” and “after the Mayor”.   But this is a pity.  For example, the poorest segment of Bristol’s population had the highest expectations of improvement from the introduction of the mayoral system in 2012, and were most disappointed by what it actually delivered in 2014 (which may have contributed to Ferguson’s defeat).  It would be interesting to know where they stand now. 

In any case one thing is clear – mayor Marvin Rees can take little comfort in the results.  Nor can supporters of the mayoral system. The system appears to have performed less well than hoped, particularly under the current mayor, as reflected in a decline in leadership ratings on every single measure between 2014 and 2018. 

Comment by Dr Thom Oliver: Co-investigator Bristol Civic Leadership Project / Dr Thom Oliver (@Thomoli) / Twitter / Bristol Civic Leadership Project | A collaborative project on change in local governance 16 February 2021:

Thanks for taking the time to look at the Bristol Civic Leadership data, I thought I would send across a few brief comments on your analysis, which I hope should explain the logics of decisions to present what we did, and the limitations of our data.

In terms of length of report, we were limited by our own individual capacity (the Bristol Civic Leadership Project takes up a small/almost voluntary amount of our ‘day jobs’ as it were), Robin Hambleton is retired, David has a senior academic role at University of Bristol with a substantial teaching load, and I am personally limited given my role is research support as opposed to a 100% academic role. We were limited by time, capacity, and the desire of Policy Bristol to produce a ‘short report’ in line with their other outputs. We still have a desire to do something more extensive, but in the academic world, our institutional push would be more towards academic peer reviewed papers, as opposed to policy briefings.

In terms of the presentation of data comparing 2012 to 2018 as opposed to all three data phases. This is on two distinct logics, firstly the BCLP aims to consider the mayoral model, not the individuals within the role of mayor. Our questions and data are depersonalised, and the original frame of the research was to compare the pre governance under the leadership model, alongside the governance after the introduction of the mayoral model. The second reason for not comparing 2014 to 2018, is that it would be statistically incorrect to draw conclusions from such comparisons. Most shifts are not statistically significant (i.e. we cannot say with enough certainty that the respective changes in scores either up or down, could be explained or attributable to a singular known reason, the other variance, which we could not account for, or attribute could be the main factor in informing the change). 

Further to this there is a particular challenge that in each cycle we are not comparing like to like between the sample populations (although Bristol City Council do very well to balance the sample), again there is variance there which given the small sample meant we couldn’t draw ‘fair’ conclusions on between data cycles.

There will be further outputs from this data, based on current data and the continuation of questions in the Bristol City Council ‘Quality of Life Survey’ and ‘Citizens Panel’, all the data is available online here: Find Consultations – Bristol – Citizen Space searching ‘Civic Leadership’. So, I would encourage people to look at the data themselves.

Overall, though, and without any statistical power, there was a definite tailing off between both mayors across several variables, and this was anticipated, given the distance from the original point of contrast, non-mayor versus mayor, and a predictable withering on the basis that original expectations of what the mayoral model could do, being met or not.


Author: Matthew Montagu-Pollock

I settled in Bristol in 2011 after 23 years in Asia, of which 16 were in the Philippines, mostly working as an investigative journalist in the Euromoney group.

Leave a Reply

Your email address will not be published. Required fields are marked *