BSIPs show how the Government can Save 15% of Civil Service Costs
Did you see that Rachel Reeves has announced that she will cut the cost of the civil service by 15%?
It was one of a spate of cost-control projects that were miraculously quantified just days before the Office for Budget Responsibility had to score their latest update to the budget.
As someone whose first job included writing National Express’s annual strategic plan, I’ve had similar late nights in the run-up to the Group financial review. It’s amazing how inventive one can be with numbers when they keep refusing to add up.
One of the National Express strategic plans I worked on. Every number 102% robust.
The problem, of course, is that having got through the Group / OBR review, the numbers actually have to be delivered.
So how is Rachel Reeves actually going to cut the cost of the Civil Service?
Well, I have an idea - and it is inspired by Bus Service Improvement Plans (BSIPs).
I’ll get onto the solution to Rachel’s problems at the end, but first let’s talk about BSIPs (I know you want to).
Happy Birthday, National Bus Strategy!
Back in March 2021, the National Bus Strategy was a bold attempt to reimagine local bus services. Officially a Department for Transport publication, it was widely rumoured to have been sexed up by Andrew Gilligan, Boris Johnson’s Number 10-based transport advisor. As well as rewriting the document overnight, Andrew Gilligan attended all the major meetings, including down to local authority level. I suspect he’d treat an accusation of micromanagement as a compliment, but I’ve personally spoken to junior officers at local authorities who - literally - found themselves on the phone to 10 Downing Street.
Once published, the National Bus Strategy dramatically reversed previous Government policy towards buses. In place of the austerity and fragmentation of the 2010s, local authorities and bus operators were to form close, integrated partnerships. The strategy was over-the-top; determining everything from how timetables should be presented to the principles of network design.
To bring all this to fruition, Local Transport Authorities (LTAs) were challenged to dream big, submitting Bus Service Improvement Plans (BSIPs) accompanied by funding bids, with the promise of transforming bus travel into a more attractive, sustainable option.
The original BSIP guidance was nothing short of breathtaking in its ambition. LTAs were compelled to craft visionary plans featuring higher frequencies, integrated ticketing, and community-centred improvements. The document repeatedly asserted the importance of scale and ambition. For example:
In Bus Service Improvement Plans, we expect to see plans for bus lanes on any roads where there is a frequent bus service, congestion, and physical space to install one. Bus lanes should be full-time and as continuous as possible.
The problem was that there was never any real indication of whether money would be available, or how much. The National Bus Strategy promised £3 billion of funding, but it didn’t seem to actually… well… exist. Moreover, for the scale of ambition, even £3 billion (vast though that would have been as a one-off bus investment), would not have been enough. Transport Authorities were all told to write breathtakingly ambitious plans and then some of them would be funded. Back at the time, I wrote on this blog:
One has that niggling worry that none of it will come to anything. A bit like if an estate agent knocked on your door and promised that your house would be worth £5 million if you just refurbished every room: if true, you’d totally be up for all the effort involved in doing the refurb job, but are you certain the £5 million will actually materialise if you put in the work?
Because the guidance covered literally every aspect of running a bus service, it felt like a lottery to local areas. Everyone had to describe transformational improvements in every area but no-one would be funded for all of it. There was no mechanism to enable prioritisation. As a result, a local authority that really wanted to trial demand-responsive buses might find that it was the bus station element of their bid that was funded, while someone who desperately needed a new bus station might find they got bus lanes. With dozens of local authorities submitting dozens of schemes each, the ‘matrix’ of schemes from which DfT had to choose was vast. From the perspective of a local area, the outcome felt somewhat random.
The Perils of Rushed Planning
Bent Flyvbjerg is an academic and author who specialises in why projects frequently go over budget. In How Big Things Get Done he tells us that waste in projects is often driven by insufficient time in the planning phase. This was a significant danger for BSIPs, as they required local authorities to spin on a dime and suddenly focus on ambitious growth after a decade dominated by cuts. Most LTAs did not have a shelf full of carefully worked-up expansion plans, as only a year earlier, no-one had been talking about expansion.
On this blog previously, I’ve described how Chiltern Railways’ Oxford-Bicester project (in which I was involved) was delivered on-time and on-budget. One of the reasons for this is that Chiltern Railways had known since 2002 that it had signed a franchise agreement with a significant capital investment commitment, and so had spent around a decade quietly optioneering. It hadn’t cost much money but different options had been worked up, tested, rejected and refined. By the time came to create a Transport and Works Act order, the plans were well thought through.
Contrast this with LTAs in 2021 who suddenly had to leap into big investments without the luxury of time or flexibility. The BSIP guidance required an immediate turnaround for highly complex plans. Those local authorities with clear visions could use the funds to implement those visions but many didn’t have clear visions (why would they?).
Inevitably, virtually all local authorities rushed to hire consultants to develop the plans. So it simply became a contest for who could hire the brightest consultant.
In some ways, it didn’t matter.
The BSIP guidance was ludicrously ambitious, describing exactly what Andrew Gilligan the Government wanted from an ideal bus service, right down to requiring that buses should have Wifi. With that level of prescription, the contest was simply who could write it up most neatly. My favourite bit of the guidance was that “letter suffix routes should be reduced”.
In some ways, it was the worst of all devolution worlds.
There are benefits to locally-owned plans, as they can be tailored and reactive to community needs. There are benefits to central plans, which can be efficient and rolled out consistently. This was an unhappy middle, in which the contest set by the Government was largely how to replay their own guidance back to them.
Managing to plans
The other problem was how to identify success. In the absence of a baseline, once a plan exists, then the only thing that the Government can do is monitor compliance with the plan. Are you doing what you said you’d do?
That’s always a dangerous place to be. One of my favourite quotes is from Dwight Eisenhower, who said: "Plans are useless but planning is essential".
What he means is that the process of creating plans is an essential thinking technique, but as soon as the plan exists, it is out-of-date. I recently met someone who used to be a manager at Avanti West Coast, and I asked him how Avanti had ended up in such a sticky place. Now, he's only one person, so he may be wrong, but he told me that the issue was that the new management team had arrived clutching the Franchise Agreement (which is, in effect, a plan) and had implemented that plan rigidly - long beyond the point when it should have been obvious that it was not right.
How to measure success
This problem was harder for BSIPs, as measuring success for the BSIPs is incredibly hard. In writing this post, I was rather hoping to be able to give you a report card on how BSIPs are doing. But I can’t.
A big problem for assessing the results of BSIPs was that this all happened in 2021, when the bus industry was in a state of Covidy crisis. The Government was providing short-term revenue support, but there was a widespread fear that this would dry up before passengers returned. This did, indeed, happen, so £1 billion of the promised BSIP funding was diverted from delivering transformational improvements to emergency funding for keeping existing services going, given the post-Covid funding cliff edge that was emerging. Even that wasn’t enough to protect networks in every area, however, so you had the slightly bonkers outcome that in some regions existing services were being scrapped (due to lack of funding) while new services were being created (thanks to ring-fenced funding).
This makes assessing the outcomes of BSIPs as a policy seriously tough. At the same time as BSIPs came from nowhere, we had:
Covid recovery taking place
Route restructures as operators adjust networks to post-Covid demand
The inflation shock impacting consumer behaviour (and bus input costs)
The £2 (now £3) price cap
Trying to isolate the impact of BSIPs in all of this is virtually impossible. It’s like trying to find the egg in a baked cake.
Indeed, given that BSIPs turned up right in the middle of the Covid pandemic there was, in effect, no baseline against which to measure success. The DfT’s evaluation document identifies four key metrics against which success will be measured:
Bus customer satisfaction
Journey time
Punctuality
Reliability
But every single one of these will have been mangled by the impact of the pandemic (customer satisfaction by the demographic changes to the customer base and the other three by changed commuting habits), making it virtually impossible to know whether any result is a good one or not.
As I’m sure you know, the gold standard for research is a randomised controlled trial. This is how medical firms assess the impact of drugs. The key principles are that they recruit patients in whom nothing else is expected to change. They then give one group the drug but not the other group. Then they compare the results in the two groups.
On this basis, a decent test for BSIPs would be to choose a time in which nothing else significant was happening in the bus industry, fund some local authorities to make certain interventions, not fund other similar authorities and compare the results.
In an ideal world, the period from 2010 would have seen a whole bunch of different interventions funded and tested so that, come the national roll out of BSIPs, an evidence base existed on what worked. It’s hard to emphasise the extent to which that is not what happened. As a result, evaluation becomes a fiendishly complex task.
One clear metric
I would also argue that this list of evaluation topics is overly complex anyway.
Journey times and customer satisfaction (etc) are all good, important things - but wouldn’t it be simpler just to focus on one clear metric? That metric could be usage. It’s surely hard to argue that a bus network which is growing isn’t succeeding. And that one that is shrinking isn’t one that is failing. Growth is the Government’s sole success metric for the economy: it’s good enough for a bus network.
It also creates the opportunity for local areas to experiment in different ways to reach that outcome. I’ll be honest: I can’t visualise how you get to growth without punctuality, reliability (etc) but maybe someone can? And if they can, don’t create metrics that invalidate their imagination.
But maybe there’s a reason why they didn’t use passenger numbers. Because when I tried, it was far harder than I had hoped it would be. For this post, I decided to use some case-studies, and asked the local authorities for their passenger journey stats. They were all completely different to the DfT-collected bus journey stats for the same counties. I tried to figure out why - and failed. I asked a couple of local authorities - but the answers didn’t really help. Is it the way cross-boundary journeys are treated? Who knows.
Here are some examples of the data from the regions I looked at. First, here’s Cornwall:
And here’s Derbyshire:
As you can see, just when we most need accurate data, the two sources wildly diverge. Why is it? Who knows? Is it to do with how cross-border journeys are counted? Either way, it means that we can’t use passenger volumes as an indicator of success.
The results On the ground
The National Bus Strategy promised £3 billion. As expected, that never materialised.
But big money did emerge.
DfT has awarded £1.35bn under the BSIP scheme (including BSIP+) (nominal). Given that the annual total net local bus funding pre-Covid was only £0.3bn (excluding London), this was a massive funding boost.
However, as expected, the scale of allocations varied significantly across areas, from just over £2 million to just under £165 million. These funds were officially allocated to enable local authorities to implement their transformational visions, but in a lot of areas the original allocation of funding was actually soaked up by the need to keep the existing network solvent, given the reduction in demand post-Covid (especially from concessionary pass-holders, who were a disproportionate share of the bus customer base. Given that this money would have come from the Government, in effect, the Government was providing Government money to replace Government money). Moreover, local authorities were told to carry on paying concessionary fare payments to local operators at pre-Covid levels, even though the passengers weren’t travelling. This was a bit of a cheat as the order came from Government but the cost was borne by councils. The areas with smaller allocations were forced to regard them as, in effect, an extension of the emergency Covid funding support.
So let’s have a look at some of the areas with larger allocations.
Norfolk
Norfolk was one of the big winners from the BSIP lottery, with an allocation of roughly £50 million (£30.9m for capital, £18.6m for revenue interventions). For a rural county, that means per capita funding of £54: massive in bus world.
Reading Norfolk’s 25,000 word Bus Service Improvement Plan, one of the things that immediately stands out is that they’ve got clear objectives. The plan is designed to achieve:
· Enhanced Accessibility
· Reliability and Punctuality
· Integrated Transport Network
· Environmental Sustainability
· Community engagement and feedback
In some ways, this is a bit motherhood and apple pie, but - nevertheless - five objectives is a lot better than fifty, which is where this kind of plan can frequently end up. To assess against these, they’ve got 12 measurable targets.
There’s also a focus on growth as a metric of success. This is also helpful. However you define a ‘good’ bus service, if people don’t wish to use it, you’ve probably got it wrong.
Both in the document (and talking to various people), you get a sense that Norfolk’s plan is a genuine partnership: and that’s tough in an environment without a single dominant operator.
Because Norfolk’s plan runs from 2022 to 2025 and is majority anchored in capital investment, it’s hard to know how they’re doing in terms of results, as we’re still in the investment phase. It’s especially hard to know given that none of the numbers seem to add up. According to their own figures, they’re now at 102% of post-Covid demand, but according to the DfT county stats, they’re at 93%. That’s a big difference and despite staring at it for a disproportionately long time, I can’t figure out why.
One of the things I like most about the Norfolk plan is that it’s actually not too rigid and prescriptive. It describes outcomes, a partnership process for making decisions and a budget. It doesn’t micro-list individual projects. That suggests the partnership board have retained flexibility to respond.
Good.
Derbyshire
The comparatively lean 18,000 words of Derbyshire’s BSIP present a more mixed picture. With a £47 million allocation - approximately £58.50 per capita (wow!) - the county’s plan is a grab-bag of initiatives: £15m targeting network pinch points, £8m for serving key attractors, £5m for ticketing officers, £4m for additional bus services (including demand-responsive transport), and another £4m for transport hubs - plus dozens more. Less than 10% is new bus services, and. significant proportion of this is for demand-responsive services, which I’m pretty cynical about (you can see my thoughts on the MK Connect service here).
The document has 26 targets that it is looking to achieve.
While there are encouraging signs - such as patronage growth beyond what Covid recovery alone could explain and strong customer satisfaction ratings (83% satisfaction, ranking above the national average) - the lack of a cohesive, clearly defined strategy is worrying. The sheer variety of schemes raises the question: can a lump-sum investment in so many disparate areas really translate into a long-term improvement in bus services?
CORNWALL
Cornwall is the ultimate example of “pot culture”, the British Government’s tendency to fund endless micro initiatives as opposed to providing consistent, predictable long-term funding.
Cornwall Council have done an amazing job soaking up different pots but it speaks to the chaos of the previous regime that there were so many different pots emerging at different times.
More or less simultaneously, Cornwall won “Superbus” funding of £23.5 million (which was a low fares pilot funded for four years) and BSIP funding of £13.3 million. A year later, they received £1.3 million of ZEBRA 2 (Zero Emission Bus Regional Areas 2) funding. All these pots have cost Cornwall taxpayers £650,000 in consultants’ fees (and, for clarity, I’m not saying that they shouldn’t have paid for the consultants: it was the right thing to do, given how many pots were available at the same time - but it’s far from ideal from a taxpayer perspective. I’m - fairly obviously - not against public bodies using consultants but winning funding pots is not a productive use).
One of the strength of Cornwall is that they did have a clear vision. They wanted to create a single Transport for Cornwall brand with a single Transport for Cornwall network. Indeed, they were already in progress with this before any of this money emerged. They had recently replaced all of their tendered services with a single mega-contract awarded to the Go-Ahead group (with independent operators subcontracted via Go Ahead), which operated half the network. In 2021, this was live - but facing significant teething troubles, which I wrote about on this blog. £37 million was always going to help fix some of the issues, and the Transport for Cornwall network is now a lot more integrated than it was.
But services that were initially puffed up as part of Superbus have started to thin out again as the fares rise, and it’s not obvious that - given Transport for Cornwall did have a clear strategy - £37 million couldn’t have been spent more effectively if they’d been able to spend it on what they wanted to spend it on, as opposed to attempting to engineer the prescribed pots into their own local priorities.
The national position
Overall, it’s hard to see that BSIPs have made a significant impact on ridership to date. The shape of decline and revival of bus patronage across the UK appears to be similar. There are local variations but the pattern seen in England outside London, Scotland and Wales (neither of which invested in BSIPs) are similar. There is not a pattern by which the locations that benefited from phase 1 BSIP funding stand out as now having significantly larger growth.
A Summary of BSIPs to date
The key conclusions from the BSIP experiment:
An exceptionally large amount - by historical standards - was invested in the bus industry
I can’t tell you if it has been well used
I can’t tell you what it’s achieved
But I can tell you that it was planned in a rush
And involved a lot of work
BSIPs revealed that while the ambition of the National Bus Strategy was admirable, the execution was hampered by unrealistic expectations and the way the exercise was carried out. Specifically:
LA’s were not ready to produce plans at a moment’s notice (why would they be?), so they end up being rushed, and not produced by local people who understand the needs of the community (and for whom councils will have had to pay more for lower quality as a result of the sudden rush on demand for consultants)
Lack of a stable baseline for the ‘do nothing’ scenario made costing of plans and the identification of what might be needed impossible
A lack of reliable and consistent data makes it very hard to know what is actually being achieved in terms of outcomes
How to Save 15% of civil Service costs
So let’s go right back to the beginning, what’s this got to do with cutting civil service costs?
Well, I’ve just spent 3,000 words describing in detail all of the work involved in making BSIPs happen. Consultants hired, bids written, bids evaluated, bids monitored, more bids submitted, more bids evaluated, more bids monitored, evaluation criteria written, evaluations written: it’s been a big job involving civil servants, officials and consultants all round the country.
How about we just… didn’t do any of this.
For real transformation, we need a shift towards devolved, long-term funding. Local authorities and operators require ongoing, predictable funding that allows them to adapt and innovate over time: not just a one-off windfall that forces them into inflexible plans. I wrote recently about how this is done in other countries. As a reminder, in Switzerland, there’s a dedicated pot of taxes that goes into a dedicated transport infrastructure fund. It all… just… happens. Similarly, local transport in France is funded through a dedicated employer levy. Yes, it can vary a little bit as firms are set up and shut down but, largely, transport authorities can predict with a reasonable degree of accuracy what funding they’re going to get in ten years’ time.
And this isn’t just transport. The crazy, labour-intensive swirling mass of bids and pots is how it also works in education, health and other parts of the public sector (remember Rishi Sunak announcing a central Government pot for school chess boards?).
If key long-term public services had certainty of funding, leaders could plan, and unit costs would be lower. It would also avoid the situation we’ve had recently when a gradual famine of bus funding was suddenly replaced by an unexpected feast. Does this graph look like the route to better bus services to you?
So perhaps the real opportunity isn’t another round of bidding guidance or a new name for the next funding pot, it’s stepping off the funding merry-go-round altogether. If we replaced short-term competitive pots with long-term, devolved funding, we could deliver better services and shrink the bureaucracy required to manage them.
We’re seeing the start of this with the expansion of City Region Sustainable Transport Settlements (CRSTSs) but it’s early and it’s still counter-cultural for the British state. But Rachel Reeves has now set a target, and this can be part of the answer. Scrap the cycles, trust local delivery and we might just hit that 15% saving without cutting a single frontline service - and allow public servants to focus on making things better in the long-run, not simply moving this year’s budget around.
