In this Data Insights series we’re diving deep into data from over 145,000 supply jobs booked through our platform to reveal data-driven best practices, trends, and insights that empower you to provide and run a first-rate supply service to support your schools. Stay tuned for actionable analyses and strategies drawn from real-world booking data and set your organisation up for success on supply.

Listen to a podcast version of this blog here.

Welcome back to our Data Insights series. In previous posts, we’ve explored patterns in booking lead times and worker utilisation. Today, we’ll take you through our analysis of worker sharing patterns across Scottish council supply pools.

Specifically, we wanted to understand the extent to which supply teachers register with multiple councils when geography makes this feasible. What we uncovered was something we’re calling “phantom capacity” – a situation where councils believe they have more available workers than they actually do because they lack visibility into where else their registered workers might be committed.

Our analysis focused on five councils within a 30-mile radius in Scotland – let’s call them Council A, Council B, Council C, Council D, and Council E. Because these councils all use our platform for 100% of their schools’ supply staffing needs, we had unique visibility into both who is registered in each workforce pool and how those pools overlap.

What we found was striking: 282 workers are registered across 2 or more of these councils, representing 9% of the total regional workforce in this area.

The strongest sharing partnerships revealed the commutable geography:

• 91 workers are registered in both Council C and Council D
• 55 workers are shared between Council B and Council D
• Multiple other meaningful overlaps exist across the region

This in itself isn’t a surprise, it makes sense that workers would maximise their opportunities by registering in all the supply pools they can reach.

When we dug deeper into which workers were multi-registering though, we discovered something particularly interesting: STEM specialists (Physics, Chemistry, Biology, Science, and Maths) are 3 times more likely to register across multiple councils compared to their non-STEM colleagues.

Looking at the STEM workforce specifically across these five councils:

• Council A: 131 STEM workers, 16% shared with other councils
• Council B: 40 STEM workers, 38% shared with other councils
• Council C: 54 STEM workers, 24% shared with other councils
• Council D: 65 STEM workers, 23% shared with other councils
• Council E: 19 STEM workers, 16% shared with other councils

Why would STEM specialists be so much more likely to multi-register? To understand this, we needed to look at how STEM work is actually structured.

When we compared STEM bookings to non-STEM bookings across our platform, a striking pattern emerged: STEM bookings are overwhelmingly long-term.

42% of STEM jobs are 21+ days compared to just 18% for non-STEM roles. Because these long-term contracts average over 6 months each, they represent 98% of total STEM working days – meaning short-term cover (1-5 days), despite being 39% of job postings, accounts for just 2% of actual work volume.

This is a fundamentally different market structure. While most supply teaching operates on daily or weekly cover, STEM has organized itself around extended contracts – maternity leaves, long-term sick cover, sabbaticals.

We can speculate about what’s driving this pattern, though we’re seeing the results rather than the intentions:

From the school side: Headteachers, aware of the well-documented STEM shortage, may avoid posting short-term STEM roles. “There’s no way I’m getting a Physics teacher for 2 days” becomes a self-fulfilling assumption, so they opt for general cover instead when a STEM teacher is sick for a day or two.

From the worker side: STEM specialists, knowing they’re in demand, may prioritize long-term contracts over day-to-day work. If you’re a Physics teacher and you know schools struggle to fill Physics roles, you can afford to wait for a stable 6-month contract rather than committing to unpredictable daily supply work.

The result: A market that has self-organised around long-term placements, with both schools and workers adapting their behaviour to the shortage. And if most STEM work is in long-term contracts, and these contracts are competitive, then registering with multiple councils makes perfect sense – you’re maximising your chances of landing one of these sought-after long-term roles.

This could explain why STEM specialists multi-register at 3x the non-STEM rate. It’s rational behaviour responding to market structure.

Here’s where the implications become significant. Each council operates as if they have an exclusive pool, but our data shows they don’t, and they have no visibility into this reality.

Consider Council B: “We have 40 STEM workers registered. That should be adequate for our needs.”

But 15 of those workers (38%) are also registered with neighbouring councils. When one of those shared workers accepts an 8-month Biology maternity cover in Council A, Council B’s effective pool has just become 39 and they don’t know it.

This is phantom capacity. Councils believe they have more available STEM workers than they actually do at any given moment. Workers registered in multiple pools can only work in one place at a time, but without shared visibility into commitments across boundaries, councils operate as if their pools are exclusive.

Let’s look at a concrete example. In academic year 2024-25, Council A ran a pre-planning process where they identified long-term staffing needs before the academic year and pre-allocated workers to schools. They successfully assigned 81 long-term STEM roles to 60 individual workers (average duration: 198 days – over 6 months), achieving a 92% fill rate for these pre-planned assignments.

Given that 16% of Council A’s STEM pool is shared with neighbouring councils, it’s statistically likely that around 10 of those 60 workers are also registered elsewhere. Once they accept these 6-month assignments, they’re unavailable to the other councils for a full school year, but those councils don’t know this has happened. They’ll only discover their shared workers aren’t available when they try to book them for roles and receive no response.

This regional market has probably always existed, but it was completely invisible until we had the data to reveal it.

We can see these patterns only because councils in this region use a unified platform that captures both supply (who’s in the workforce pools) and demand (what jobs are posted and filled) across administrative boundaries. Without this infrastructure, each council operates with only their local view. They know who’s registered with them. They don’t know:

• Who else those workers are registered with
• When those workers become unavailable because they’ve taken assignments elsewhere
• Whether they’re competing with neighboring councils for the same workers during peak demand periods

It revealed that councils have been operating on incomplete information about their actual workforce capacity.

The Scotland data is unique because councils provide all supply through managed pools, and multiple councils in this region use the same platform. This gives us complete visibility into both sides of the market.

England’s market operates very differently. Supply is provided predominantly by agencies in a highly fragmented, competitive marketplace. But workers in England almost certainly engage in the same rational behaviour: registering with multiple agencies to maximise their opportunities.

Without shared infrastructure, however, there’s no way to see the extent of this overlap. When a Multi-Academy Trust establishes a Preferred Supplier List with 5 agencies, a reasonable assumption might be that this provides 5x coverage. But do those agencies actually provide additive capacity, or is there significant overlap in who they can deliver?

A platform that captured both supply and demand across agencies could potentially reveal:

• The extent of worker overlap across agency panels
• How often multiple agencies submit the same candidate for the same role
• Whether a ‘diverse’ PSL actually provides the coverage schools assume it does
• True capacity vs perceived capacity across an agency network

We can’t prove this happens in England but the fundamental behavior (workers maximising opportunities through multi-registration) is likely very similar. The question is whether organisations operating in fragmented markets have visibility into these patterns, or whether they’re managing based on assumptions about capacity that may not reflect reality.

We’ve literally just discovered these patterns ourselves, and councils haven’t had the opportunity to act on this information yet. But the discovery raises important questions:

• How do councils plan workforce capacity when they don’t know which workers are shared or when they become unavailable due to commitments elsewhere?
• Are there STEM workers who would take short-term work if it was offered, but schools don’t post it because they assume it won’t fill?
• When multiple councils experience simultaneous demand (say, September planning phases), are they unknowingly competing for the same workers?

Markets organise themselves based on worker behaviour and economic incentives. Digital infrastructure that spans boundaries can reveal market dynamics that individual organisations, operating in isolation, simply cannot see. And once you can see the system as it actually operates rather than as you assume it operates, you can begin to ask better questions about how it works and where the real constraints lie.

In our ongoing Data Insights series, we’ll continue exploring what becomes visible when you have comprehensive data infrastructure.

Our ongoing Data Insights series will continue to offer valuable perspectives and practical guidance to help schools master absence management. Keep an eye on our blog to stay informed!