Research and analysis

BDUK Hubs Evaluation Plan

Published 15 August 2025

1. Introduction

Building Digital UK (BDUK) commissioned Belmana, Hatch and Winning Moves to conduct an evaluation of its Hub intervention. The Hub intervention invests in connecting remote public buildings, such as schools, general practice (GP) surgeries and libraries, to gigabit-capable broadband. The initial Hub intervention was delivered through BDUK’s Rural Gigabit Connectivity programme (RGC) which ran between 2019 and 2022. The next phase of the Hubs, which are referred to as GigaHubs, are being delivered as part of Project Gigabit. The planning for this iteration of the Hubs intervention began in 2021 and are being delivered between 2022 and 2026. The evaluation is being conducted between 2022-27. 

The first phase of the evaluation concluded in December 2023, with the publication of the report BDUK Rural Gigabit Connectivity Hubs Evaluation, focusing on the Hubs delivered as part of the Rural Gigabit Connectivity Programme. This evaluation plan covers subsequent phases for the evaluation from 2024 to 2027. The plan covers the tasks that evaluate:

• Phase 2: processes associated with the Hubs delivered through Project Gigabit and the impacts they achieve

• Phase 3: early impacts of the GigaHubs and medium-term effects of the RGC Hubs, looking at effects up to 2024/25

• Phase 4: the further impacts of the Hubs until 2027, returning to Hubs multiple years after their connections

The plan sets out the theory of change, the data to be used in the evaluation, the methods used, and the timings of the key activities involved in delivering the study.

1.1 Programme description

BDUK Hubs fund gigabit broadband connection to a public building that is remote from the fast broadband network. In rural areas, the costs of connecting a primary school or a GP surgery is likely to be very high, and not affordable for a supplier purely based on providing a broadband service to the building for the usual contract period. The BDUK funding subsidises the connection. This study focuses on two programmes that use the intervention.

The Hub intervention was launched in 2019 as part of BDUK’s Rural Gigabit Connectivity (RGC) programme as a ‘demand-side’ intervention, meeting the demand for speeds faster than 1000 Mbps in a public building. It funded connections until 2022 to public buildings in remote (usually) rural areas. Investing in public sector building upgrades was a feature of the earlier Local Full Fibre Network, whose Public Sector Building Upgrade (PSBU) projects provided a basis for the intervention by focussing on rural areas who had less than 100Mbps available in the surrounding postcode.

In 2021, the Hub intervention was incorporated into Project Gigabit, a government programme to support the rollout of gigabit broadband to hard-to-reach areas. Hubs delivered under Project Gigabit are referred to as GigaHubs. It funds gigabit-capable connections and associated infrastructure in areas which are unlikely to receive this purely through commercial investments. Alongside the funding to connect public sector buildings, the wider programme was similar to RGC in offering subsidy for the delivery of gigabit-capable broadband in specific areas and delivering voucher schemes for rural regions. Project Gigabit undertook projects that were unfeasible under RGC due to the time required to deliver and the scale of the project. These barriers were particularly prominent for Government bodies with a large number of dispersed premises.

1.2 Investing to improve rural connectivity through Hubs

The Hub model was piloted in BDUK’s Local Full Fibre Network (LFFN) programme, and then further tested by the RGC programme. The intervention involves identifying eligible public sector buildings which meet qualifying criteria set by BDUK. These criteria consider several factors, primarily targeting rural areas with a low threshold internet speed – sites must have a current download speed below 100Mbps (with sites below 50Mbps particularly prioritised). In addition, eligibility involves considering the funding required, State Aid compliance, and other planned broadband infrastructure investments (commercial or otherwise) in the area.  Further criteria include value for money, and deliverability within the timescales of the programme. 

BDUK worked with strategic partners, including the Department for Education (DfE) and the Department of Health and Social Care (DHSC), as well as devolved administrations and local authorities. These partners could lead projects, essentially identifying and aggregating eligible rural public sector buildings to act as Hubs and then procure the works to provide broadband infrastructure. 

Procurement would be by the lead organisation, with the option to use procurement systems set up by BDUK to go to the market. BDUK has set up frameworks to run competitions for works using standardised documentation and processes and then reaching all the main relevant telecom infrastructure suppliers. This enabled lead organisations to easily tender for connections.

Identifying eligible public buildings has been key. Organisations leading projects put in place processes to identify public buildings – such as schools, GP surgeries, and libraries, many of which experienced poor connectivity and could be remote from the network. The lists that were compiled were tested in terms of the criteria set by BDUK, with delivery commencing in 2019 under the RGC programme and then continued under Project Gigabit in 2021. The RGC Hubs were evaluated, reporting first findings in 2023.^{[footnote 1]}

1.3 Objectives of this report

The objective of this report is to update the plan for the evaluation of the Hub intervention. It aims to provide details about how the rest of the study will meet the objectives of the evaluation. The replanning of the evaluation follows a review in 2024 as the extent and pace of delivery of the Hubs become clear. 

There are likely to be fewer Hubs than envisaged at the outset of the evaluation in 2022, and the period over which delivery is to take place and the years for which impact analysis is needed is lengthier. In addition, the concept of connecting a public building to gigabit-capable broadband has matured. The Hub intervention increasingly operates alongside other BDUK products and programmes. The Gigabit Infrastructure Subsidy (GIS) is the most significant. GIS is a government programme aimed to support the build of gigabit-capable broadband infrastructure in the areas of the UK which are unlikely to attract commercial investment for gigabit-capable broadband infrastructure within a reasonable timeframe, if at all, referred to as the final 20% (f20). It contracts with suppliers covering the whole network in an area, with the advantage for BDUK of greater control over the delivery of connectivity to premises beyond the Hub. The GIS has then provided an avenue for suppliers the opportunity to aggregate demand and connect premises at scale; it has also reduced the number of public buildings eligible for the Hubs. The replan has sought to integrate these wider context and drivers. 

2. Theory of Change

This evaluation focuses on the BDUK Hub intervention, delivered in two programmes. The Rural Gigabit Connectivity programme (RGC) piloted the use of Hubs, and the evaluation began as this delivery was nearing completion targeting connecting around 800 public building sites to gigabit-capable broadband. As part of Project Gigabit, GigaHubs were introduced in April 2021. The aim was to connect up eligible rural public sector buildings, in order to improve public services. 

This section first describes the benefits framework BDUK applies. It then describes the logic model underpinning Hubs. The chapter then describes the theory of change underpinning how the intervention, once delivered, has impacts. It then concludes by listing the evaluation objectives.

2.1 BDUK benefits framework

BDUK have developed a benefits framework which provides a single, coherent plan of the benefits which each BDUK Programme aims to achieve, and how these can be measured. The following sections give a summary of the benefits expected to be realised from the Hubs intervention, which have been identified in the BDUK benefits framework.

Direct benefits

The BDUK benefits framework identifies enabling public sector efficiency as a benefit area. The Hub intervention will make delivery of public services more efficient with better resource use in in management, administration and through the reallocation of tasks and work hours due to digitalisation, increase adoption of new technologies specifically in the services delivered at Hubs, improve quality of services and offer new services at Hubs. 

It also identifies stimulating the broadband market as a benefit area. The Hub intervention will achieve this by making infrastructure available in very hard to reach areas. For suppliers, this reduces barriers to entry and providing consumers more choice. 

Indirect benefits

Indirect benefits are derived from the take up and use of gigabit-capable connectivity. The Hub intervention will contribute to reducing the digital divide and providing public value by providing public access to reliable broadband by increasing uptake of gigabit connectivity on upgraded infrastructure in areas that may have taken longer to be connected without government subsidy. I it will promote uptake of business and residential connections in the surrounding area as the Hub brings infrastructure closer to the properties reducing cost of connection. Residential uptake of broadband can affect subjective wellbeing, community wellbeing and the value of housing in remote areas. Intermediate outcomes include access to training and entertainment and enabling access to services online.

It contributes to driving growth in the economy by reducing connection costs, which would increase the take up of connections by businesses. The Hub intervention will provide gigabit infrastructure to small and medium enterprises (SMEs) located and operating in rural and remote areas. Better connectivity may bring better employment; connectivity helps workers to become more efficient and utilise time and skills. It would allow the adoption of new technologies and the associated skills.

There is also an indirect effect on enabling public sector efficiency because the Hub intervention will provide or enable access to reliable broadband for users of public services. The gigabit connectivity provides access, participation in and use of digital public services.

Finally, the BDUK benefits framework identifies reducing impact on the environment as a benefit area. Technologies which enable individuals and firms to be more energy efficient, using online meetings rather than travel. Commuting reduced lessens Co2 and Nox emissions.

There are different dimensions to how the benefits realised through connecting a public building in a remote area are articulated.

There are benefits at the strategic partner level because the Hubs are used to aggregate demand across a number of public buildings, providing consistency and future-proofing the connectivity across the estate, with potential cost saving and ability to plan strategically for use of it in future.

The Hub itself and its service users experience benefits because the Hubs are able to deliver public services using digital technologies. The outcomes from the subsequent service changes and then wider effects can then be modelled. This could be simultaneously as premises are passed at the time of the delivery of the Hub or later given the existing infrastructure should reduce cost to build in the surrounding area.

At a wider spatial level, it is expected that effects will be experienced around the public building from further non-subsidised build effects.

2.2 Spatial dimensions of the Hubs

There is a spatial dimension to benefits. The Hub intervention connects eligible public sector buildings (such as schools, GP surgeries, and libraries). While this will improve public service at the Hub, it also makes it cheaper for nearby premises to connect to gigabit capability and draw in suppliers for the wider network. The benefits will spread out to benefit not only the Hub but also the wider community through improved availability, leading to increased take up and eventual use of gigabit connectivity. 

In promoting the Hub intervention to key stakeholders (devolved local and national government, suppliers etc), BDUK sought to visualise this staged and geographical model, shown in the following figure.

The figure shows a series of nested concentric circles. In the centre of the model, the provision of a gigabit-capable connection to a public building enhances a public service.  If done as intended, the delivery of a connection to the public building will also allow for near immediate delivery of services to the neighbouring premises with little or no additional cost, and certainly at a cost that the market provision would meet. This is depicted as a wider concentric circle surrounding the first.

Surrounding this is a circle that represents the effect on business and residential voucher interventions. With supplier presence established through the Hub, they can stimulate business and residential demand to establish gigabit in the still uncommercial areas around the Hub, perhaps using the voucher product. Hence there is expected to be some overlap between Hubs and vouchers in the same intervention area. The final outer circle represents some of the sparsely distributed premises that remain out of reach.

The selection of the Hub is crucial, as this mix of public service and local outcomes meant navigating the needs of public buildings and the infrastructure plans of suppliers, as well as interacting with other broadband and digital policies. Sponsor bodies were identified, such as the Department for Education (DfE), the Scottish Government (SG), and local authorities (LAs) to list potential Hubs.

2.3 Hub intervention: developing a theory of change

The Hub theory of change has been developed alongside other BDUK products. It analyses the intervention’s approach to delivering benefits, identifying demand for gigabit connectivity to meet an area’s connectivity needs with this centring on the needs of eligible public sector buildings (such as schools, GP surgeries, and libraries). 

These outcomes and impacts are achieved through different mechanisms involving the Hub, suppliers, and other government bodies. First, the Hubs provide funding for connecting public sector buildings delivering essential services in the community. 

Making broadband available in connected Hubs increases availability in surrounding areas. This leads to a cluster of benefits ‘stimulating the broadband market’ in the figure. The connecting of nearby residential and business properties is likely to be enabled, as indicated in the model of spatial dimension also. It also as some indicative timelines for this. However, the progress over time of broadband take up in nearby areas will differ from area to area. 

2.4 Inputs

BDUK funding

£110m of the total funding allocated to Project Gigabit Programme has been made available for Hubs, although some flexibility is built in to this so that funds may be diverted if required. The GigaHubs are not expected to be continued, and Hub activity is due to reach completion in 2026, so no further funding is due to be allocated beyond the current allocation. 

BDUK contracting / management expertise

BDUK drives local procurement behaviour through a combination of advice, support, documentation, grant agreements, and the assurance process. The assurance checkpoint process, managed by BDUK, provides an opportunity to confirm that commercial principles are enacted in procurement documents and resulting contracts. BDUK also support partners in other government departments and in local authorities through providing funding and assuring delivery against pre-agreed milestones. The BDUK funding is only to cover the capital elements of connectivity upgrades.

BDUK data

This includes knowledge, lists, data, and information on the public premises in need of connectivity (from LA/Gov etc). i.e. the need/demand being aggregated. 

Eligibility criteria

BDUK have a specific set of criteria by which public sector buildings are identified as eligible for inclusion in a Hubs project, based on premises level supplier build information, and our assessments of probable commerciality based on premise density where supplier plans are not complete. This aligns GigaHubs eligibility with GIS procurements and voucher criteria, and includes areas that would otherwise be left behind, delivered later, or more expensive.

Wider government department resources

Time and strategic resources are required from partner departments such as the Department for Education (DfE) and the Department of Health and Social Care (DHSC) to deliver the projects in which they take the lead. DfE is BDUK’s leading partner in a project that at time of writing has 787 Hubs in scope in England. NHS Scotland have also delivered 5 Hub sites in the Western Isles, Highlands, and in Orkney. Other government departments own the process of procurements and are responsible for the contract management of their projects. 

Local authority funding and resources

Three GigaHubs projects are being led by local authorities, in Dorset, Oxford, and Leicestershire. As with the DfE and NHS Scotland projects, the local authorities are responsible for resourcing the contract management of their projects. This usually involves appointing a project champion within the authority, to take the lead on procurement, management, and promotion of the project. 

Commercial supplier finance and resources

Project Gigabit aims through the provision of subsidy to attract commercial investment and incentivise broadband suppliers to expand into areas that are otherwise unviable, particularly that face significant geographical and logistical challenges that increase the cost and complexity of deploying gigabit-capable broadband. Whilst the Hub intervention is not directly expected to result in commercial build, it brings additional build to remote areas through supplier activities connecting public buildings.

Central government learning

The Hub intervention has been tested through the Local Full Fibre Network (LFFN) programme and the Rural Gigabit Connectivity (RGC) programme. The lessons learnt from these projects are used to inform how the GigaHubs are managed.

2.5 Activities

There are multiple activities which need to be completed to ensure the delivery of the GigaHub intervention. These are highlighted below.  

Project initiation

GigaHub projects begin with building collaboration with partners who will be leading on the projects. Project initiation activities including establishing the lead organisation and project champion and developing a working partnership with stakeholders, designating the roles and responsibilities within the leading organisation, and establishing the parameters of the collaboration. 

Identification of eligible premises

This involves developing the scope of the project through building a site list which identifies potential Hubs that meet the criteria and developing a procurement strategy. This includes activities such as the Open Market Review (OMR) that BDUK conducts every four months that identify suppliers’ commercial plans for build. Specific BDUK eligibility includes identifying a site classed as a public sector building and performing a public function, a rural location defined using an agreed standard of measures depending on the devolved nation, an existing broadband speed of less than 100Mbps, being out of scope of commercial build for a gigabit-capable network, and out of scope of other interventions in the area.

Scoping the project

This involves establishing and developing the project scope using the Business Case Toolkit, establishing the strategic, economic, commercial, financial and management case for the project. 

Procurement of contract

This involves identifying a suitable procurement option using the Dynamic Purchasing System (DPS), issuing an invitation to tender in the preferred market route, evaluating responses in accordance with scoring criteria, selecting the preferred bidder and notifying successful (and unsuccessful) bidders. Approval of the grant offer award is given upon receipt of assurances and completion of submission. It brings with it a commitment to execution of build. 

Submission

Using criteria and navigate assurance process steps (including site level dialogue) to turn this into a final list of sites for procurement/contract. This includes preparing the documentation required for assurance and liaising with BDUK Local Delivery Partners or Managers to agree the best solution for the specific contract. 

Management of project delivery

This is overseen by the lead body who is responsible for the management of the supplier. Part of this activity is reporting back to BDUK through fulfilment of project management information templates and compliance with financial deadlines. 

Benefits tracking and monitoring

BDUK require projects to collect additional management information from the supplier to assist with benefits monitoring, which feed into evaluation. Typically, this reporting should be submitted every quarter for three years following completion of the project build. 

2.6 Outputs

The key outputs that the GigaHubs are aiming to achieve are outlined here.   

Delivery of gigabit-capable broadband to remote or rural areas

Sites that have been identified as being outside of commercial build plans will receive gigabit-capable broadband.  

Expansion of infrastructure network

Infrastructure will be made available in areas that were previously deemed unviable by commercial suppliers, making the area more commercially attractive for supplier-led build out in the future. 

Provision of gigabit-capable infrastructure to public sector sites in remote communities

Public sector sites in remote and rural areas will have access to improved connectivity services that will deliver considerable benefits in terms of enabling service transformation. 

Government learning

Extensive and proportionate evaluation of the Hubs provides the opportunity for BDUK to improve future government investment

2.7 Outcomes

Improved long-term reliability of infrastructure at the site

The improved infrastructure will provide a long-term solution to providing reliable broadband at the site. This will provide access to infrastructure that can support use through multiple devices simultaneously at a site with multiple users and will ensure that a reliable connection is available for as long as the site is within use. 

Additionality

Premises outside of the scope of suppliers’ commercial build plans will receive gigabit-capable broadband infrastructure at the identified premise. This results in the increased availability of faster, more reliable broadband at the premise. 

Take up of gigabit contracts

Improved infrastructure will create new services to be offered at the public sector sites and the opportunity for public sector organisations to take up contracts that include gigabit-capable broadband at those sites. 

Adoption of public sector digitisation strategies

Public sector buildings start to adopt the digitisation strategies published by their government departments. These are specific to the building in question and will differ from site to site but broadly include the digitisation of processes that have been identified and evidenced as being suitable for digitisation, the increased use of digital technology at the sites, the adoption of new and emerging technologies, and their integration into service delivery. 

Expansion of digital strategy and procurement

As digitisation expands, emerging technologies and updated devices are procured by government departments and local authorities. 

Application of government learning

The learning achieved during the outputs stage can be used internally to improve BDUK products, but is also shared across government departments through published evaluation research reports. 

Citizen use of digitised services

Service users become accustomed to accessing information, transactional services, and other digital services online and become more comfortable with their use. As digital services become more integrated to provision, the ‘digital first’ approach suggested by Government Digital Services meets service users’ needs whilst also making more efficient use of public sector resources. 

2.8 Impacts

Longer term expected impacts of the GigaHubs programme are centred on promoting public sector efficiency as the buildings are giving access to fast, reliable broadband infrastructure. Project Gigabit more broadly aims to improve public sector efficiency in a quantifiable manner, based on the logic that better internet connectivity can lead to efficiencies in service delivery within the public sector. These sectors include, but are not limited to, health services, educational institutions, leisure facilities, local authority functions, criminal justice services, and energy and transport sectors. This broad variety of public services are able to make better use of digitised public services and utilise sector-specific enhancements to increase their efficiency. 

2.9 Provision of public services 

Improved resource management   

Resources in this context refers to the objects, devices, and platforms used to deliver public services, but it also refers to the people who use them and the time spent doing so. So whilst this refers to the use of (for example) cloud storage for easier access to important files, it also refers to the automation of certain tasks that allow for time and effort to be redirected elsewhere. This is particularly true of tasks that can be automated, such as minute-taking.  

Improved efficiency in service provision  

It is supposed that improving the management of public sector resources will allow departments to spend more time and effort in carrying out tasks that are necessary to the provision of services. There is also an assumption that the digitisation of tasks will implicitly speed them up and make them more efficient. The extent to which this happens across public sector departments will depend on many factors, especially given the wide range of services that are being discussed. For example, the provision of driving licenses will differ to the delivery of health care. Many of the tasks that receive this benefit have already been digitised but it should be noted that it is not necessarily universal. A good example of this is using digital technology to triage patients in secondary care, and attending routine appointments with patients with chronic illnesses. The evidence suggests that these are processes that can improve the efficiency of healthcare provision but there are still many details missing from the wider evidence as to how this can be implemented more broadly.  

Improved capacity to integrate emerging technology 

Evaluation evidence suggests that public sector organisations will redirect finances into digital technology once they begin to see the benefits of it. As the adoption of digital technology becomes embedded into the public sector, evidence suggests that organisations will also be able to integrate technology more quickly as it emerges. This means that the benefits of improvements to automation, cloud computing, and machine learning can be more quickly realised as organisations are better prepared to integrate them into their ways of working. 

2.10 Access to public services 

Faster access to public services and information 

The purpose of the digitisation project is to make public services more accessible to a wider range of people whilst minimising the impact those who deliver the services. The foundations of this have been laid by Government Digital Services who have enabled information and statistics to be presented digitally in a way that promotes engagement and access across all sectors of society. This progress is continuing with the development of app-based services, such as apps developed by the healthcare services, and the digitisation of DVLA services. Such developments improve efficiency by making information available thus reducing the need for contact with service providers. The project is in the process of being developed, through further services as the UK Gov App which collates all of the services a person might need in one place.  

Data-driven (personalised) decision making 

The development of app-based services allows for a holistic approach to service provision by government, driven by data. This in turn allows for data-driven decision making that makes service provision and intervention more personalised and tailored to individual needs. This is particularly useful in the case of complex needs where public sector services are accessed across a range of government departments. 

Improved mental health and an increase in healthy lifestyles

Increased connectivity creates opportunities for increased and more discrete opportunities for mental health support, which in turn improves wellbeing, creates greater outcomes for service users, and reduces pressure on mental health services. Digitisation also creates opportunities for healthier lifestyles through allowing access to online wellbeing resources such as exercise classes, healthy eating information, and social and cultural activities. This also creates better health and wellbeing outcomes for users, reducing pressure on health and social care services. 

Reduced impact of public services on the environment

Environmental benefits from the integration of digital technologies include increased energy efficiency at public sector sites and increased use of cloud computing rather than using physical servers which are more power intensive.

Improved future government programmes

The application of government learning from the Hubs intervention can be used to improve future programmes. 

2.11 Disbenefits

Low public trust and engagement

It should not be assumed that the digitisation is desired or welcomed by those who access services. At present, there is very little evidence to help government services understand the perspective of service users, and it is unclear whether digitised services improve the user experience of those accessing the services. It is unclear whether the digitisation project will result in greater satisfaction with services or promote engagement with essential services.  

Limited scalability of infrastructure  

This disbenefit refers to the significant limitations placed on public sector bodies in their capacity to be reactive to emerging technology, and to adopt new forms of digitisation once their reliability has been proved. Barriers in this area are chiefly public sector reliance on legacy technology, which is well established within the literature in this area, but it also refers to scalability in terms of skills, capabilities, and willingness in public sector employees. The digital skills gap that exists in the UK is felt keenly amongst many public sector departments and it is noted that there is a significant lack in leadership that enables this to be resolved at present.  

Further exclusion of poorly served communities 

The development of a ‘digital first’ approach brings with it the risk of excluding citizens who struggle to access digital services. The assumption that services will be accessed and used digitally creates several layers of exclusion whereby those who lack the connectivity, devices, skills, or confidence to access services online will not be able to gain the help that they need. Research from organisations such as the Good Things Foundation and academia suggest that those at highest risk of digital exclusion are those who already experience exclusion and barriers to participation in society. People who live on low income, live with disabilities, or are isolated are among those who are more likely to be in this group, and the digitisation of public services risks exacerbating and entrenching those difficulties. 

2.12 Evaluation aims and objectives

The evaluation primarily focuses on the effects of the Hub and what impacts are attributable to the policy. As the Hub intervention is within a wider programme of demand-side and supply-side interventions and supports the delivery of public services, the study has covered process aspects, understanding the scoping and delivery of the Hubs and how the Hub results in better outcomes. The research questions ask about direct (broadband) outcomes but then also their onward effects. The attribution of these to the Hubs investment and the logic of the policy is crucial to help identify how to test this. Identifying the logic of the policy is crucial to be able to attribute these effects to the Hub investments.

Evaluation aims and objectives

The overarching aim of the evaluation is to establish the connectivity impacts of Hubs. BDUK programmes have a standard set of overarching evaluation questions that must be assessed, in line with the Magenta Book and Green Book. Evaluation projects seek to answer or contribute to the overall understanding of these questions: 

• What are the outcomes of the Hub intervention across RGC and in Project Gigabit? 

• What has changed in individuals’/organisations’ behaviour for these outcomes to come about? 

• How effective and efficient has the delivery of the programme been? 

• Was the investment cost-effective? 

• What can we learn to improve future policy designs and implementation? 

These questions are answered in relation to a BDUK benefits realisation framework, which includes three priorities: stimulating the broadband market, enabling public sector efficiencies, and ensuring an estimation of the full impact.  

For the first priority, the evaluation aims to look at how the Hubs influence the market through stimulating investment, increasing competition, and addressing the market failure to provide gigabit-capable connectivity. The second priority is to be considered in terms of establishing measurable benefits to Hub site beneficiaries, especially where these can be monetised and/or estimated robustly. These differ by the type of Hubs, and this study focuses on the effects on educational services as schools are the main type of Hub site.  The final priority, that the evaluation gives a full picture of impacts, is in the context of indirect local area benefits, and this requires a baseline estimate of both intervention and suitable counterfactual areas.

3. Data and data collection

This section looks at the data that is being compiled to explore the Hubs connections. Various secondary datasets are used, and data has been collected using surveys and interviews. This chapter describes the data, first defining the unit of analysis, and then looking at the individual datasets or sources from which data is collected. The main secondary data is the management data and various published broadband and socioeconomic data. Much of this data is geographically defined, so that spatial analysis is possible in the impact analysis. Quantitative data is complemented by qualitative sources of evidence, with the study conducting surveys, using the results of other BDUK surveys and interviewing stakeholders associated with the Hub intervention. 

3.1 Overview

The Hubs will deliver different impacts at various levels and the evaluation’s evidence gathering is tailored to these. This chapter looks at the data that will be compiled or collected, and the methods to be used for each. Table 3.1 links each of these to the types of benefits outlined in the previous chapter. The approach in the evaluation has been to use: 

Management information. The delivery of the Hubs intervention has associated management information about the public building connection and some of the associated policies such as broadband vouchers and the Gigabit Infrastructure Subsidy (GIS). 

Secondary datasets. A strand of quantitative analysis focuses on the broadband performance at the Hubs and in the area around the Hub. Using management information and secondary data, a linked dataset is compiled for small areas in the UK covering both Hub areas and comparable unsupported ones.  

Survey data. The public sector efficiency benefits and broadband market effects will be complemented by qualitative evidence. The surveys of the Hubs and in-depth interviews will look at the public sector efficiency and will provide insight about the broadband market. 

In-depth interviews. Some benefits will be enumerated using in-depth interviews. Management information will be used to identify the sample to be surveyed or contacted for interviews. The evidence gathering will then explore with participants the benefits that are observed in reducing the digital divide, driving economic growth and reducing the environmental impacts. Questions in surveys will ask whether internet access means increased business opportunities, or reduced travel. The design of the survey will seek to explore effects that are additional, attributable to the Hubs.

The following table sets out how data will be collected to meet each pillar of the benefits framework.   

Table: data collection and benefit analysis

x = used to gather data

Benefit	MI	Secondary data	Surveys	In-depth interviews
Enabling public sector efficiency	x	x	x
Reducing the digital divide	x		x	x
Driving growth in the economy	x		x	x
Stimulating the broadband market	x	x		x
Reducing impact on the environment	x		x	x

The analysis will have different units of analysis. The wider broadband network is also affected by Hubs, as the investments bring broadband into the network around a Hub. For these area effects, the level of analysis is at the Output Area (OA). There are about 240,000 OAs in the UK and each has about 150 properties, large enough to provide useful statistics but small enough to represent some of the geographical differences of local areas. Overall, units of analysis are: 

Direct effects at the Hub level: The direct effects of the access to gigabit infrastructure will be the take up of fast broadband and then its use in the delivery of public services at the Hub, at more efficiently operating at the Hub and of providing new services. The relevant unit of analysis is the premise and the activities taking place in the setting, and data can be collected about the Hub, the postcode it is in and the output area that contains the Hub 

Indirect impacts on businesses and households: The Hub connection brings the gigabit-capable network to areas that otherwise are quite distant from fast broadband. This can then lead to the take up and use of broadband. The unit of analysis for this effect is nearby households and businesses. 

Indirect impacts on broadband use in nearby areas: The infrastructure investments affect the broadband network near to the Hub and effects are likely to be on small areas. The unit of analysis is output areas, and analysis can take into account location such as distance from a Hub. 

3.2 Management data

The BDUK Management Information (MI) provides information on progress in delivery of the Hub intervention and supports the assessment of impact by identifying the beneficiaries of the investments in public buildings’ broadband. Also, the study benefits from access to wider BDUK MI, covering the vouchers issues to businesses and residences to connect premises to gigabit broadband. This section describes the management data. 

Hubs monitoring data

The Hubs were delivered through two programmes, with each having a set of projects that delivered the Hub connection. The two waves of Hub projects are: 

Rural Gigabit Connectivity Hubs: In early 2019, an initial set of Hubs targeted areas that were deemed to be the hardest to connect, where speed was lower than 50Mbps (later 100Mbps) and that were classified as rural. This led to around 1,100 sites being identified. There were ten projects that were initiated through the sponsor bodies.  

Project Gigabit GigaHubs: In late 2021, BDUK began to approach suitable organisations about applications for GigaHub funding and the processes involved to make the case to BDUK for funding. Individual projects were led by six central and local government organisations scoped around 2,500 candidate sites, and about half of these were thought to be eligible. Projects began delivering Hubs in March 2022, expecting to deliver connections until March 2026. 

The two largest projects are led by DfE with a project in each of the RGC and GigaHub waves of funding. Over the two waves, there have been four devolved administration projects, and some medium sized ones delivered by local authorities. Often the projects involve partnerships, such as local authority led projects involving National Health Service (NHS) sites. For projects, there were leads in the funded delivery organisation and then sponsor leads within BDUK (called Local Delivery Leads), facilitating the allocation of the RGC/GigaHub funding to the sponsored bodies who could then procure the connections. 

3.3 Data collection using surveys

There are two surveys planned during the evaluation, one conducted by the study team and the second being led by BDUK collaborating with DfE: 

Baseline and follow-up survey of libraries: in 2024/25, a baseline survey of libraries distant from the gigabit-capable network will be conducted. For libraries that are connected to the network through a BDUK investment, a follow-up survey will be conducted around six months after the connection. A sample of the other libraries would also have a follow up survey in 2026/27.  

Schools survey: A survey of the schools benefitting from the Hub intervention undertaken twice, once before the connection and the second after the connection. The survey results will be integrated into this evaluation. 

The Hubs surveys ask those running the (broadband) services in the Hub, asking about the school or library broadband connectivity and the uses made as the Hub provides public services, either library or educational.  

Surveying libraries about connectivity

The libraries survey is another component of the evaluation of BDUK’s GigaHubs programme, focusing on assessing the impact of gigabit-capable broadband connectivity on public library services. The survey is designed to capture both baseline conditions prior to broadband upgrade and the subsequent changes following the intervention.  

Purpose of the libraries survey

The primary objectives of the survey are to  

1. Assess the availability, reliability, and performance of broadband connectivity in libraries before and after intervention. 

2. Explore how improved connectivity affects the range and quality of digital and non-digital services provided by libraries, including support for education, job search etc 

3. Examine the ways in which library patrons and staff utilise broadband-enabled services, including frequency and purpose of use. 

4. Investigate the broader impacts of enhanced library services on community engagement, digital inclusion, and accessibility. 

Design of the libraries survey

The libraries survey will be conducted in two waves to capture data at different stages of broadband connectivity: 

Baseline Survey (2024/25): Conducted before libraries are connected to gigabit-capable broadband, this wave will collect data on existing connectivity levels, digital infrastructure, and the services currently offered. 

Follow-Up Survey (2025/26–2026/27): Conducted six months after the broadband upgrade, this wave will gather data on changes in connectivity, service delivery, and community outcomes. It will include libraries that received BDUK-funded upgrades and a sample of non-beneficiary libraries for comparative analysis. 

Both waves will use consistent survey questions to enable before-and-after comparisons while tailoring specific questions to reflect the stage of connectivity. 

Survey questionnaire

The design of the questionnaire will be informed by a review of library surveys conducted in the UK and US and discussions/feedback from library experts in DCMS. The survey will explore several critical themes to understand the impact of broadband connectivity, including: 

• connectivity and infrastructure, covering current broadband performance, reliability, and speed; existing IT infrastructure and digital resources in the library 
• digital Services and usage, asking about types of digital services offered, such as e-books, online learning, or public access terminals, frequency and purpose of broadband usage by library patrons, and barriers to accessing and utilizing broadband-enabled services
• impact of broadband, covering perceived changes in library operations and service delivery, and outcomes for library staff and patrons, including improved access to information, education, and other public services 

*respondents will be asked if they are willing to participate in follow-up qualitative interviews to provide more in-depth insights

To ensure clarity and relevance, the survey questionnaire will undergo cognitive testing with a small sample of libraries. This process will help refine questions, response options, and the overall structure before full deployment. 

Sampling approach

The libraries survey will focus on public libraries across England, with a particular emphasis on those within the scope of BDUK’s GigaHubs programme. This includes libraries that will receive gigabit-capable broadband upgrades funded through the programme, as well as a comparable sample of non-beneficiary libraries to provide a control group for analysis. 

A flexible sampling approach will be adopted which includes libraries connected through other BDUK funding routes (expected to be mainly GIS). The sample will be adjusted as more information about connected libraries becomes available.  

The sampling frame will be derived from a comprehensive list of UK libraries provided by DCMS, which includes: 

• library names, types, addresses, and postcodes
• broadband connectivity status and estimated cost of connection
• operational details such as staffing, opening hours, and co-location with other public services 

Libraries that could be included in the sample will be identified using the following sampling criteria: 

• connection costs: prioritising rural and remote libraries, especially those in the 20-25% highest cost areas for broadband connectivity
• commercial viability: focusing on libraries categorized as “uncommercial,” “hold-up,” or “beyond value for money” in the BDUK data

Our initial analysis has identified 801 libraries that meet these criteria (see table below).  In addition, there are some libraries in the scope of a Hub project that do not fall in the 25% or 20% most expensive areas. These have been added to the survey sample so that the before/after surveys cover all Hubs (these libraries are shown in parentheses). Overall, a sample of 740-873 libraries will be contacted for the survey. 

Table: sampling libraries for the baseline survey

BDUK Category	Libraries in England and cost to connect
	All libraries	Costliest 25%	Costliest 20%
Commercial	2,071	190 (+69)	82 (+74)
Uncommercial	418	340	331 (+2)
Hold-up	367	270 (+3)	243 (+7)
Total	2,857	801 (+72)	657 (+83)
Sample		873	740

The survey will aim to achieve the following sample sizes across two waves: 

• baseline survey (2024/25): 100 completed surveys targeting libraries without gigabit-capable broadband
• follow-up survey (2025/26–2026/27): 100 completed surveys, split between libraries upgraded through BDUK funding and non-beneficiary libraries

Survey implementation

Recruitment for the libraries survey will use the administrative data provided by DCMS. This data includes a contact email address for each library, enabling outreach via email. Key steps in the recruitment process include: 

• email invitations explaining the purpose of the survey and its importance in evaluating the impact of broadband improvements
• follow-up reminders to non-respondents to maximise response rates 
• clear communication about survey confidentiality and the option for respondents to participate in follow-up qualitative interviews 

The surveys will be administered online using ID survey to facilitate efficient data collection. Respondents will access the survey via a secure link provided in the recruitment email.  

During the fieldwork period, response rates will be monitored to identify and address potential gaps in representation. Steps to ensure data quality include: 

• reviewing responses for completeness and consistency 
• applying robust data security measures to protect respondents’ information

3.4 Stakeholder depth interviews

The evaluation evidence gathering will involve around 75 depth interviews conducted to deliver qualitative findings about Hub impacts. The estimate is for around 10 interviews for the process review and then further interviews for impact assessment. The expectation is that there will be 20 interviews at Hubs, 10 interviews with those involved with projects (such as BDUK policy and Local Delivery Leads, project leads in LAs and government departments, and suppliers) around 35-45 business or residential interviews.  

In-depth interviews of staff at Hubs, suppliers, households, and businesses 

Those working at schools, surgeries etc connected using the Hub intervention are to be interviewed about their use of the internet, and the changes attributable to the faster connection. The study will include interviews of broadband suppliers, policy leads and – following the online survey of voucher beneficiaries – in-depth interviews of a sample of the residents of these connected properties. 

Recruitment into the depth interviews is challenging. The collection of qualitative evidence on an area level, allowing some understanding of the delivery of the Hubs from the local perspective, has proven difficult during earlier stages. This is primarily because the delivery was complex, with differing models across the projects and the staged approach meaning individuals’ involvement can be brief. For example, many of the stakeholders that were involved in the early stages of the process were unavailable to provide evidence about the planning stages or early delivery. There has also been a set of parallel policy initiatives – such as support for technology adoption alongside investment in the connection – that mean stakeholders may not view the Hubs individually and find it difficult to describe attributable impacts. 

The main purpose of these interviews is to gather more in-depth insights and views on benefits of the Hub upgrade on the public services delivered in rural areas and the effects on the wider community. Recruiting individuals for the interviews will involve a number of steps:

s1.Residential interviewees. For the BDUK evaluation of the Gigabit Infrastructure Subsidy programme, gigabit voucher recipients are being surveyed and, working with Ipsos who are leading the evaluation, voucher beneficiaries near to Hubs will be identified and, at the end of the survey asked if they would be willing to participate in a follow up conversation about their upgrade experience, in greater depth. The respondents that say they would be happy to be contacted for this purpose form the sample for the qualitative interviews with residents.  

1. Library Hubs. The library survey will be conducted in early 2025 for a baseline and the survey will ask if recontact is acceptable for an in-depth interview. When the follow up survey is conducted (sometime during 2025-26), a further sample will also be recruited for depth interview. 

2. Other types of Hubs, project leads and suppliers. Contact will be through the projects, such as DfE Schools Project enabling contact with primary schools connected; or LA-led projects introducing to recycling centres, health centres. Suppliers would also be contacted through the projects. 

3. Community level contacts. Depth interviews of community stakeholders will be recruited directly by the team. In the areas with local authority projects, a set of interviews of parish level officials and councillors would be conducted. Follow on connections from these and other local contacts will be compiled. 

Analysis will look across the different interviews and combine with other contextual evidence. This will enable assessment of the effect on the public building and the services provided but then analyse wider community level effects including the widening of access to fast broadband, or the benefits due to the changes in the services provided at the Hubs. Evidence sources would include such as websites and online documents and reports. For example, the papers associated with local council discussions about broadband have proved valuable, as well as the press releases, newsletters and marketing of suppliers, Hubs and community groups were used in earlier fieldwork.

Broadband outcome data: coverage and performance

Ofcom publishes the Connected Nations reports on the UK’s communications infrastructure, focusing on coverage and performance of fixed broadband and mobile networks. This report uses the annual report data from Ofcom Connected Nations from 2018 up to 2022. The annual reports track progress, focusing on the availability and performance of fixed broadband and mobile networks, and this is published at a detailed geographical level (postcodes and output areas) which can then be averaged or aggregated at the higher OA level.  

The data enables year-on-year comparisons of the UK’s communications infrastructure in terms of both the availability of broadband at different speed levels and performance (data on change in average speeds). This dataset is used to obtain the main outcome variables of the analysis.  

Connected Nations relies on data provided by suppliers and cannot be guaranteed to provide full coverage of all fixed networks. For this study, the data from the annual report is used as it is a final version for the year, filling gaps in the interim updates. The annual report data was used for each year from 2018 to 2022. 

Most of the variables on the fixed coverage focus on the percentage of premises that meet certain speed availability cut-offs or could be connected to specific technologies. A key feature of Connected Nations is the geographical detail that is available, with data files generally being by postcode or output area, and – for many variables – available for both levels of geography. 

Looking over the annual reports, there is a variation in the variables covered often reflecting technological changes: the data about the availability of full fibre is up to 2019/20 year and Gigabit is recorded in the most recent years. The technologies classified to full fibre and Gigabit – while broadly similar – diverged significantly for Virgin Media and there is a discontinuity in the time series about coverage associated for areas supplied by the provider. Broadband coverage statistics cover almost all UK output areas. An issue with postcode level datasets is the limited availability of counts of properties that are consistent with the shares data that is published. There are also a significant number of postcodes for which data is unavailable because the count of properties is too low. 

Alongside the coverage statistics, Connected Nations has measures of broadband performance. The variables available for fixed performance are the minimum, average and maximum download speed, as well as the data usage, and the number of connections. The data was collected in May of 2018/2019/2021/2022/2023 and in June 2024. 

3.5 BDUK datasets

Building Digital UK (BDUK) provides a range of data for the study, in addition to the intervention data related to the Hubs supported (and described earlier in this chapter). This section describes the data, one about a parallel product relevant to the logic of the Hubs – broadband vouchers – and other data BDUK uses to capture connectivity in the UK. 

BDUK vouchers, voucher projects, and premises passed

It was expected that once a rural public building has been connected, further connections may sometimes need subsidy. A voucher is the BDUK contribution to a business or residential contribution so that the supplier can then provide a fast broadband service on a commercial basis. BDUK voucher management data – and there have been different waves of the voucher support - covering the timing, supplier, value and location of each incident of voucher support was provided. There have been several hundred thousand applications for a voucher, with many properties then being provided a voucher.

Where the voucher was a project, the individual premises passed by the projects’ investments was also made available. A voucher project might connect several households or businesses but then the infrastructure would pass other premises that would now be able to take up a gigabit service. The data provided these potential additional premises that could take up gigabit broadband, and for each the unique property reference number (UPRN) was usually indicated.

Data is compiled to track both the number of premises in an output area connected using a voucher and the number passed in each output area annually by the projects. 

Open Market Review data

The BDUK Open Market Review (OMR) is a process to assess the availability of broadband infrastructure across the UK, particularly in areas that may require public intervention to improve connectivity. A key aspect of the Review is data collection. The OMR gathers information from broadband providers about their current and planned network coverage (usually over the next three years). Telecom companies, local authorities, and other stakeholders are invited to submit data on their coverage to help ensure public funding is targeted effectively. 

The review helps to identify areas where no commercial broadband investment is planned, meaning they may need government funding to improve digital infrastructure. Data is collected at a premise level, using the UPRN, asking about broadband technology, the speed available, and what commercial broadband investment is planned. The data informs gaps in coverage and the need for government funding to improve digital infrastructure. OMR is a preliminary step before government-funded broadband expansion projects, such as Project Gigabit. It is being used extensively in the GIS programme. 

This evaluation can use this new data source and a phase assessing the data and then scoping its use will be key to the next stages of the study. This will both provide a more complete picture, especially in areas under-served for high-speed broadband which would include areas with Hubs. The Data Protection Impact Assessment and data sharing agreement has been amended to reflect this activity.

Telephone exchanges and their served areas

Each property is connected to an exchange, through cabinets and inter-cabinet infrastructure. While data on the property to exchange is not made available, there are postcode to exchange lookups, such as the SamKnows database, and then the premise to exchange connection has been modelled by BDUK to give the geography of the telecom network. BDUK have made this available to the study. 

There are 5,583 exchanges and each serves around 5,000 premises. The Hubs can be matched to the exchange serving them and the same exchange can serve multiple Hubs. While there has been substantial investment in rolling out the connections to exchanges by fibre-optic, the exact times and extension over time is not available for this analysis. So, rather than characterising the exchanges, the analysis has constructed variables based only on which exchange serves a Hub. The exchanges that connect to the Hubs can be identified. Also, each output area that was served entirely by a single exchange can be determined and the ones where the exchange which connected to a Hub tagged. These output areas would have the same broadband infrastructure as the Hub up to the exchange. This could characterise an output area as similar to the Hub areas. They therefore provide a pool from which a counterfactual can be chosen which shares a major technology characteristic with the Hub. Also, if the Hub connection involved some improvement to the exchange, then the properties in the OAs that use the exchange may also benefit. 

A further variable can be derived from the exchange dataset. The postcode of the exchange meant that the distance from each OA to the exchange could be estimated. This would also correlate with the costs of connecting the properties.  

F-score modelling and assessing connection costs

The commercial viability of UK locations is modelled by BDUK. The F-score Model provides an estimated relative cost to install fibre to the premise. This has been indexed for the model, but the value is proportionate to cost, so that low values reflect premises that can be connected at modest costs, while higher values indicate the opposite. The premises have further been characterised in terms of their viability, primarily to differentiate those that can be connected commercially in that the market would reach on its own without public subsidy. The categories are commercial, meaning premises would be connected on a commercial basis and the three categories where public support may be needed. This commercial/non-commercial distinction approximates to the 20% of premises referred to as being F20 premises in earlier related work. 

The BDUK cost modelling focuses on properties in the UK, estimating the build out costs from a fibre hub, i.e. an exchange or some other location with fibre connectivity. The 2022 model uses a September 2021 dataset about the fibre connectivity status of cabinets and then models costs to connect each property from there. The property list used in the modelling is epoch 91 AddressBase Premium (ABP) and, while scores are estimated at property level, analysis always aggregates into areas recognising modelling is not specifically undertaken at property level.  

Costs are estimated by multiplying the modelled building distance a measure of cost per metre providing a categorising of costs into the four subsidy categories (commercial, hold-up, uncommercial and the beyond VfM, with this the most costly). The majority of properties in the dataset are commercial with 82% in this category. 

Data to characterise Hub areas

The modelling seeks to characterise areas in terms of their business population. For this, the ONS Business Register and Employment Survey has been used to provide data on employment at LSOA level for England and Wales (via Nomis). The equivalent series for Scotland and Northern Ireland were integrated. As well as total employment by LSOA, a variable measuring employment in digital sectors was constructed. For OA level analysis, only LSOA level data is available. 

Other area characteristics controls: ONS/NRS/NISRA population density estimates at LSOAs, the 2011 Rural Urban Classification for LSOAs, Indices of Deprivation, and the Internet User Classification (IUC) from the Consumer Data Research Centre (CDRC), which allocates LSOAs to different categories based on how households interact with the internet. For OA level analysis, again only LSOA level data is available. Overall, variables derived from these public datasets are: 

Population density: The number of people per square kilometre in Lower Layer Super Output (LSOA) areas in England and Wales, Data Zones (DZ) in Scotland and Small Areas in Northern Ireland. 

2011 Rural Urban Classification for LSOAs/DZs and Small Areas: The Rural Urban Classification is produced using Census data, with the 2011 Rural Urban Classification being the latest version of the classification. The next Rural Urban Classification will be produced when the 2021 Census data has been published. We include a dummy identifying rural areas. 

Indices of Deprivation: These datasets provide a directly measured Index of Deprivation across all LSOAs in England and Wales, as at 2015-16, enabling ranking across the two countries. Equivalent analyses for Scotland and Northern Ireland have been added and analysis includes these rankings both with England and Wales and interacted with a dummy for the two Nations to allow for a different scale effect in the compilation of the indices in Scotland and Northern Ireland. 

2018 Internet User Classification (IUC): The IUC is a bespoke classification that describes how people living in different parts of Great Britain interact with the Internet. It provides coverage for Great Britain at the LSOA (for England and Wales), Datazone (for Scotland) level and Small Areas for Northern Ireland. The IUC provides 10 unique profiles of neighbourhoods based on a number of characteristics,  

Many of the Hubs are primary schools. For English schools, a public dataset provides the location and key characteristics of the schools. This is the Department for Education Find School Performance data service1. Further, the DfE Consistent Financial Reporting (CFR) framework provides a framework for English schools to collect information about their income and expenditure. Financial year data for 2017-18 to 2021-22 is used for this analysis. The variables drawn from these sources: 

Number of pupils at the school: this is used to indicate primary school size, with the Hub schools generally smaller. 

Pupils’ characteristics: the variables drawn from the dataset are the share of pupils that are eligible for free school meals and the share than receive support for special education needs. 

Expenditure on ICT learning materials: expenditure per pupil CFR code E20, ICT learning resources. 

Total expenditure per pupil: aggregating across staff, materials and consumables used in the school. 

Investment in ICT: ICT equipment in total across schools. 

The variables are linked to the school reference number, and then to the address of the school. The postcode can be linked to the Hubs database so that all English primary schools supported with Hub investments can be characterised in terms the number of students and the pupils’ characteristics.  

There are also datasets about the GP surgeries, tracking services of which some are likely to be affected if the surgery’s broadband speed improves and demonstrating whether the local area around the Hub also benefits from fast broadband. Activities such as consultations may go online, and these would be tracked by the  evaluation.

4. Evaluation methods

Impact evaluation assesses the Hub intervention’s achieved outcomes, and estimates what impacts are additional, attributable to the BDUK investments made into Hubs.  A mix of methods are employed for the evaluation. These explore the outcomes that occur after Hub investments and qualitatively and quantitatively looking at the changes observed. There are then different methods to understand to what extent these ‘gross’ outcomes would have happened without support. The complexity is that any intervention will be occurring in the context of wider changes in broadband provision and be affected by other complementary drivers for the use of a broadband connection. The plan outlines steps taken to reduce the effects of this, and to understand the impacts due to the Hub intervention. 

4.1 Overview

The plan includes quantitative analysis evaluating impacts as well as more qualitative methods. This section gives an overview of these approaches.

For quantitative analysis, an output area dataset of broadband coverage and performance is compiled, linked to other broadband outcomes such as the vouchers connected, and premises passed by gigabit-capable network. Further the dataset includes data about the performance of the public buildings in an area. There is also data about the economic and social characteristics of the areas, including their remoteness from the broadband network. The approach statistically identifies a counterfactual, which is used to estimate impacts. 

Survey evidence centres on the responses of schools and library Hubs to questions about the change in broadband, the uses made by the Hubs of the broadband in delivering services and then respondents’ views on the extent to which they attribute the changes to the Hub intervention. 

Qualitative evidence will be gained using in-depth interviews, asking more details about the impacts and explaining how they occur. While the number of interviews can be modest, there are advantages to interviewing across a wide range of stakeholders, gaining insight from the Hubs, but also those that supply broadband connections, or work across many schools and libraries, or are local residents, businesses and community leaders. The multiple sources allow analysis to triangulate across different perspectives. 

Additionally, the study is looking at the processes underpinning GigaHubs delivery and the value for money of the Hub intervention. 

For process analysis, an opportunity to study the rollout of GigaHubs is to be conducted alongside the delivery of the programme. This seeks to answer research questions especially around the selection and procurement of Hubs. The approach uses in-depth interviews similar to the qualitative evidence above with some analysis of management information. Survey evidence centres on n the extent to which they attribute the changes to the Hub intervention. 

4.2 Impact evaluation challenges

The impact evaluation methods track outcomes, establishing the changes as Hub connections takes place and then using quantitative and qualitative methods to establish what impacts are attributable to the Hub intervention. The next section describes methods to be used, both quantitative and qualitative, and how these address the challenges in estimating additional impacts.

Deadweight and identifying the counterfactual. A counterfactual provides what would happen anyway, without support. This is the deadweight of an intervention and is used to understand additional impact. However, the accuracy of any estimates crucially depends on how good the method has been at establishing the comparator.  

Parallel broadband policies. The Hubs policy is taking place in the context of wider policies to encourage gigabit broadband. Attributing impacts to the Hub intervention needs to ensure effects due to parallel policies are not inadvertently attributed to the Hub intervention. 

Parallel public services delivery policies. Similarly, the Hubs policy is taking place in the context of wider policies to encourage the adoption of digital technologies using broadband at Hubs, such as investments in educational technology at schools. Some outcomes at the Hub would occur because of the combined policies, and a focus is assessing how and whether the Hub has enabled the parallel policies’ effectiveness. 

Delivery timescales and time to impact. Some impacts associated with the Hub intervention may take a significant time to materialise. An evaluation may be too early to capture the wider uptake of broadband in a community or the further effects of the fast connection as Hubs integrate their connectivity into operations.  

Wider impacts of Hubs. The logic of the Hub intervention envisages indirect effects – premises neighbouring Hubs taking up broadband more cheaply, suppliers entering areas due to the Hub connection. These are difficult to fully assess, both because the scope can be wide and because there are difficulties finding a counterfactual. 

These challenges underpin the approaches planned for the evaluation. The next sections look at the quantitative counterfactual impact evaluation (CIE) approaches. Data about the treated areas will be compared to other comparable areas. The later sections look at the survey evidence and the approach for in-depth interviews. There are also methods described to assess value for money and to undertake the process review required for the GigaHubs programme. 

4.3 Counterfactual impact evaluation

In counterfactual impact evaluation, a small area dataset of broadband coverage and performance is compiled. The unit of analysis is mainly the ONS output area, a Census 2021 level of geography. This is linked to the Hubs list by identifying the output areas that contain at least one Hub investment, plus other broadband outcomes such as the vouchers connected, and premises passed by gigabit-capable network. There is also data about the economic and social characteristics of the areas, including their remoteness from the broadband network and variables about the public buildings in an area, whether supported by the Hub intervention or not. 

Overview of approach

Counterfactual impact evaluation explores the supported public buildings, the communities and suppliers nearby, but then also what is occurring in comparators. In this way one can compare the average performance of the treated (after the intervention) to the average performance of those untreated (in a period comparable to that of the intervention). Any difference can then be attributed to the policy but only where the evaluator can assert confidently the treated and non-treated cohorts are alike, i.e. a counterfactual group can be defined to compare the treated with.  

The method then uses propensity score matching to identify output areas that do not contain an investment for each of the areas with a Hub. Propensity score matching is a statistical technique that establishes the features of a typical supported Hub area, allocates a score to it and then finds unsupported areas with similar scores because they have similar characteristics. This depends on the variables available to match the Hub areas to unsupported areas, and whether these adequately characterise the areas in terms of what merits Hub support.  

The performance in outcome variables is then tracked over time in these two types, with any significant differences attributed to the Hubs intervention (difference-in-difference). To attribute effect, the most common quantitative approach is to construct a counterfactual, modelling what would have occurred without the support and then viewing any changes seen in the Hub area not seen in the comparable ones as due to the investment.  

Such before/after analysis reaches a reasonable level of robustness, usually classed as SMS level 2 (Standardised Maryland Scale, SMS, measures the most robust as 5 and least as 1). In addition, the dataset allows a statistically matched comparator set of unsupported areas to be identified, which raises the SMS level to two.  

Identifying the counterfactual

The first stage is to define the support, with the Hub intervention providing fast broadband to a single property. Its direct effect on the network is necessarily quite limited, and of a different magnitude to interventions such as the vouchers where hundreds of thousands of properties are directly connected through the supplier. The pathways by which the further indirect effects may then materialise would be at an area level, with the area defined as being close (within a kilometre) to the Hub and so most likely to be able to extend on the connection provided to the public building. 

The following diagram presents the counterfactual modelling employed in terms of the areas that contain Hubs and the options for the selection of a counterfactual. An explanation of the diagram is given here.

The data compiled covers all primary schools, libraries and GP surgeries, though early phases focus on schools. Over the study, knowing which areas have a public building means the evaluation can match to unsupported areas that include the types of public buildings and schools. 

Comparable OAs are then selected using propensity score matching from all OAs. This is named pool one and is represented as a large grey circle in the diagram. Within this large circle can be found the OAs that were connected to the same telephone exchange as a Hub. These are labelled as pool 2 and are represented as a smaller white circle within the grey circle, which is connected to circles representing the Hub areas and nearby areas. The areas that were near to a Hub but distant enough to be unaffected by the Hub connection are labelled pool 3. These are 4-5km from a Hub OA and are considered to be fringe areas.

A final pool 4 – the OAs that have a public building – allows analysis to match not just on the broadband infrastructure, socioeconomics, and rurality of an area, but on the fact of an area containing a public building. The nature of the Hub can also be included in matching, matching to both similar areas and similar buildings. 

A further analysis in pool 4 is possible, focusing into the impacts the postcodes that contain a Hub relative to the postcode with the public building in a matched OA. This will be possible for this pool because the postcode of the counterfactuals can be identified and there is broadband performance data for most of these postcodes. 

The pools from which counterfactuals are selected offers analysis to look at effects geographically. Analysis restricts the comparator to proximate areas, either in terms of being geographically close (in a circle 4-5km from a Hub) or in the telecom network sense, in sharing a telephone exchange with a Hub. Proximate areas can act as a good comparator because they will share many of the technical and socioeconomic characteristics of the supported areas, and therefore be more comparable.  

Pool 4 is important in restricting selection to a defined pool of areas where there are public buildings. The selection modelling only drew the counterfactual from output areas known to have a primary school, GP surgery or library.  

In the modelling to find comparable output areas, selection uses Probit model to assess what determines areas being supported on observed variables. This can capture many of the features of the supported areas that distinguish them as being different. The following table indicates the variables that are used in the RGC impact evaluation’s selection models. These are derived from the secondary datasets described in section 3. 

Table: variables used in selection models

Variable	Description	Used in models (1 – 5)
Distance to exchange	Distance from the LSOA / OA centre to the exchange serving at least 60% of properties	1 - 5
Employment (logged)	LSOA level employment for England, Wales and Scotland; NI is larger areas	1 – 5
Index of multiple deprivation	Rank for each country with interaction term for Scotland and Northern Ireland as ranks are separate for two countries	1 – 5
F20	Modelled score for the cost to connect property, averaged for OAs	1 – 5
Rurality	Binary variable based on 2011 Census ONS rural/urban classification	1 – 5
ICT employment	Share of LSOA employment that is in ICT industries	1 – 5
Population (logged)	LSOA level population estimates from 2011 Census	1 – 5
Voucher beneficiary	At least one voucher connection in area in year before	4 – 5
Change in download speed	Connected Nations area level data for annual chance in year before	3 and 5

Table: Variables used about schools where OA is in England and has a primary school  

Variable	Number of pupils in the school	Used in models (1 – 5)
Free school meals	Percentage of pupils in the school eligible for free school meals	1 – 5
SEN share	Percentage of pupils that receive support for special educational needs	1 – 5

Multiple models are estimated, using different sets of variables in the selection models and then applying to different pools of unsupported areas. The table provides details of variables, but models are broadly similar having rurality, socioeconomic characteristics and measures of the broadband infrastructure. The alternatives then add in broadband performance variables, such as the speed of broadband before the Hub investment and whether an area received voucher support before Hub investment.  

The selection modelling that restricts to areas with a public building (pool 4) can include variables about the public buildings. The table indicates the variables that can be used where the public building is a school. So, for the schools, matching would select output areas where the schools are similar in the number of pupils, or the share of pupils eligible for free school meals. 

Measuring impacts

Where selection modelling is considered robust so that counterfactuals have been identified, a difference-in-difference approach is used as the main way to estimate additional impacts. It explores whether the growth seen in supported areas and the control group (the first difference) differs significantly in the two groups (the second difference). For most estimates, the second difference will consider the changes seen in the Hub areas with the counterfactual and any difference can be both tested (showing additional impacts) and compared with the changes or level in the supported area (to indicate what the gross effects, without adjusting for deadweight, is in supported areas). 

Table: outcome measures in the impact evaluation (measures of broadband uptake)

Effect	Description	Units
Number of vouchers in OA	BDUK gigabit vouchers across schemes	Number
Value of vouchers in OA	Value of vouchers across schemes	£
Numbers of connections >30Mbps	Premises accessing fast broadband	Number of lines
Average download speed	Connected Nations speed at postcode and OA level	Mbps, logged

Table: outcome measures in the impact evaluation (measures of broadband availability )

Effect	Description	Units
Premises passed by BDUK funded project	BDUK voucher project premisses passed compiled at area level	Number
Gigabit availability	Connected Nations share of premises with gigabit available	% premises
Gigabit availability, logged	Connected Nations share of premises with gigabit available	% premises, logged
UFBB availability	Connected Nations share of premises with UFBB available	% premises
Unable to receive 30Mbps	Connected Nations share of premises without fast broadband available	% premises

The following table splits the main variables about broadband into those that measure availability, quantifying how many or what share of premises could access a broadband speed, and those that measure uptake. For uptake, the actual average download speed is a good measure as, when residents or businesses take up a gigabit connection the average speeds in their output area or postcode will rise. Availability is different in counting the properties that could potentially take up a fast broadband connection. 

The vouchers datasets also track uptakes. The compilation of the vouchers would focus on those that receive a voucher supported connection and then the dataset contains the time when this is served by a provider. 

When analysis focuses on supported areas and selects the counterfactual from the areas that have a public building, as wider set of outcome variables can be used in analysis. So, in models that use pool 4, as well as looking at broadband uptake and availability the difference in difference analysis can look at outcomes measured for the individual Hubs compared to the public building in the counterfactual. 

Table: outcome measures in the impact evaluation (measures of broadband uptake)

Effect	Description	Units
Expenditure on ICT learning	English schools financial returns to DfE for expenditure E20	£
Capital expenditure on ICT equipment	English schools financial returns to DfE for expenditure E20	£
Online consultations	NHS Digital Data Link: General Practice Appointment Data Practice level breakdown	Number

The table highlights the outcomes available about schools. Individual schools annual produce financial accounts, covering their expenditures on IT, both on ICT learning and on equipment. These have been compiled for the RGC evaluation and the performance seen in Hub schools have been compared to the counterfactual to estimate whether the Hub schools spend more than expected on these digital technologies. This can be used as an outcome. 

For those areas with GP surgeries, the NHS Digital Data GP Appointments Data provides a practice level breakdown over time. This dataset provides detailed information on the number of online consultations. Again, as the availability of broadband should enable more online consultations, there would be the possibility of using the NHS Digital Data to create an outcome variable that can estimate whether there have been additional online consultations. 

For the difference-in-difference modelling, maintaining reasonably high sample sizes is crucial for robust analysis. Further sample sizes will need to be high for the sub-samples of different types of Hubs. For the impact analysis: 

Stacked analysis will be used. The approach taken will centre analysis around the year before support, with the outcome variables then defined in terms of the year after, two years after etc. The years of data will then be stacked so that the supported are analysed across the years of support rather than analysing each year separately. 

Postcode level data about outcome will be used where appropriate. As the Hubs and public buildings generally are premises, and some outcome data is available at the postcode level, using the postcode level outcome data is possible for model 4. Postcode level data, based on a small number of premises including the public buildings, will more accurately track changes in the performance of broadband speed. 

Premise level data will be used where appropriate. Similarly, where the outcome variable is about the public building – or at least the services provide there – such as the school or GP surgery data, then this will also provide more precise outcome data. 

Analysing impacts using Hub survey evidence

Individuals managing or working in public buildings that have benefited from a Hub investment will have been surveyed before the connection (the baseline survey) and more recently, after the improved connection. Surveys would cover primary schools and libraries. The baseline surveys will focus on connectivity levels before investment; follow up surveys will be conducted after broadband has been improved. 

The questionnaire covered connectivity and use at the time and then explored the use of the internet in services provided (educational at schools and library services). Questions would look at any barriers experienced. Where a connection is known to be imminent, the questions also cover the preparations made by the Hub. 

Overall, survey analysis would cover: 

• connectivity and satisfaction with performance 

• before-after connection use of internet 

• impacts on learning and staff workloads 

• barriers to technology use

Analysis aims to determine the past position of the use of the internet and then explore the changes using both categorical and textual responses. Key to this will be that the design of the follow-up survey uses baseline questions from the earlier survey and, while record linkage over the Hubs across the two waves is not possible, the average change between the surveys offers insight about changes seen. Further, wider public surveys, such as for schools the Education Technology Survey conducted by DfE, offers some further comparable insight about broadband use. 

Library survey analysis

The analysis of the library survey will assess the impact of gigabit-capable broadband on library services and community outcomes, using both quantitative and qualitative methods. Quantitative analysis will: 

• measure changes in broadband performance, digital services, and community engagement between baseline and follow-up surveys 

• compare results between libraries benefiting from BDUK funding and non-beneficiaries to identify the programme’s effects

Qualitative analysis will examine open-ended responses to uncover themes such as connectivity challenges, benefits, and innovative broadband applications. These insights will complement the quantitative findings by providing context, including illustrative examples and success stories. 

Schools Hubs survey analysis

The analysis of the schools that were connected to broadband using the Hub intervention will be conducted by BDUK, but its analysis will be part of the impact evaluation of the Hub intervention.  the impact of gigabit-capable broadband on school services and community outcomes, using both quantitative and qualitative methods. Quantitative analysis will: 

• measure changes in broadband performance, digital services, and community engagement between baseline and follow-up surveys

• compare results between libraries benefiting from BDUK funding and non-beneficiaries to identify the programme’s effects

Qualitative analysis will examine open-ended responses to uncover themes such as connectivity challenges, benefits, and innovative broadband applications. These insights will complement the quantitative findings by providing context, including illustrative examples and success stories. 

4.4 Analysing impacts and in-depth interviews

The aim of the in-depth qualitative interviews is to understand the processes and pathways by which Hubs deliver impacts and provide a picture of the key outcomes and impacts. Interviews cover the challenges faced at each of the stages both to draw learnings and to test the logic of impact. Various stakeholders, including local delivery leads or managers at BDUK, LAs or other project leads, suppliers and Hubs are interviewed to gather evidence on how Hub impacts occur, and the nature of the impacts seen.  

Case study approach

The fieldwork seeks to undertake case studies about areas where projects have been delivered. The case studies will involve interviews at different stages of the study and so provide a baseline of effects but then return – where this is practical – to consider medium and long-term changes. The projects are often area specific, targeting the public buildings in a local authority area, and these LAs are generally the means to structure the fieldwork. In addition, many Hubs are delivered by national projects, such as the DfE Schools project connecting primary schools all over England. Some fieldwork will target the project (such as interviews with project leads or key suppliers at a national level) but then the aim would be to cover the Hubs in the areas where case studies focus. 

The BDUK local delivery leads (LDLs) for each of the local authority projects provided a first set of interviewees. The interviews explore the role of the LDLs, aims of the project, how the project was scoped including the process to select sites, delivery and challenges faced, status of the project and anticipated benefits or early outcomes. The interviews with LDLs help gain an overarching understanding of the different delivery stages and gain insights into each of the projects. 

Further contacts may be snowballed from this initial contact. Such as the LA managers delivering the projects. Interviews can tease out the processes used in projects, synthesising across views on project aims, how hubs were selected and delivered. The interviews can take the local context into account and explore the different project stages to impact. National delivery bodies, such as Department for Education (DfE) are similarly contacts for fieldwork. 

Typically, the delivery bodies can enable contact with suppliers, Hubs and further stakeholders such as community groups involved in the Hubs project. The approach to contact suppliers was tailored by project leads identifying contacts and the best pathway to contact that individual to understand supplier experience in delivering connections. The suppliers were asked about their previous involvement with Hubs, impact of the Hubs on their activities, relationship with the LAs and Hubs, and learning from the project.  

Analysing interviews

The fieldwork will be underpinned by interview frameworks. The frameworks were designed to cover all stages of delivering impact from a Hub and then adapted to focused on a few stages depending upon who was being interviewed. Different stakeholders added to the understanding of how Hubs were delivered, and evidence was consolidated across the piece to highlight the process at each stage. The same set of frameworks were used by all researchers conducting interviews in different study areas.  

4.5 Approaches to monetising impact from the Hub intervention

The value for money assessments to be made for Hubs has to take account of challenges. The logic of funding a connection to the public building is to improve the delivery of services at the Hubs and to make fast broadband accessible to local residents and businesses near to the Hub.  

To measure the effects at public buildings, the identification and quantification of public sector outcomes is out the scope of the study. This is because of the known complexity of measuring public sector, non-market, outputs.  

Further understanding the wider indirect effects on broadband network of a connection at a remote public building is difficult. There are a number of policies that are directly connecting residential and commercial properties, there are also investments in broadband infrastructure. These BDUK policies are being evaluated separately, with evaluations of the Gigabit Voucher Scheme and the Gigabit Infrastructure Subsidy being progressed. If the counterfactual impact evaluation can extract any local effects of Hubs precisely, these studies may provide valuation methods. The approach to VfM then focuses on two main points. Firstly, it will examine the potential to look at impacts at Hubs qualitatively and then implement approaches on defined monetizable effects at Hubs, such as cost savings and or resource reallocation that can be attributed to faster broadband. Secondly, it will integrate the findings of parallel studies using surveys of residential and business voucher beneficiaries near to Hubs that will enable some analysis of what impacts the voucher beneficiaries value and the monetizable impacts. 

Approaches to monetise Hub impacts

This section focuses on the first bullet, starting from the Hub intervention theory of change and including the qualitative evidence. The following section develops the theory looking at different types of Hubs, including primary schools, GP surgeries, and library or community centres. It considers the logic of the Hub intervention applied to Hub type. 

Direct benefits from Hubs

The schools that are included in the Hub intervention are primary schools with an average pupil number of 145, and an average of £75k spent on IT. The logic for the the hub intervention is to overcome IT use barriers due to unreliable or slow broadband, and to have additional broadband speed and reliability due to the Hub investment. In schools, this thought to be detectable through the adoption of educational technology and through improved administrative efficiency.

GP surgeries and libraries are also being connected as a way of overcoming barriers to IT use through unreliable or slow broadband, and to have additional broadband speed and reliability due to the Hub investment. For GP surgeries, it is expected that these will be detected through adopting technology for treatment pathways, improved efficiency in keeping patient records, and improved communications. For libraries, it is expected that the Hub intervention will result in the increased use of IT by library patrons and the adoption of new services within the library. The extent to which these expectations are true will be met through the Hub evaluation by assessing various impacts both quantitatively through secondary data and surveys, and qualitatively through in-depth interviews.

Evaluation approaches for VfM are developed for these direct effects. Some have been used in the RGC evaluation, which estimated the additional impacts at primary schools only. These were then further corroborated by more qualitative evidence. The section then looks beyond schools, at approaches that look GP surgeries. 

At schools, qualitative evidence indicates that the provision of a gigabit-capable connection improves speeds and reliability which then enables parallel investments in equipment. These allow the adoption of educational technology in schools. The qualitative evidence in the RGC Evaluation also has evidence that technology improved administrative tasks, making school operations more efficient. This provides the first expenditure-based approach. 

Increased IT expenditure

The table highlights the dimensions for evaluating impacts. First, it shows the scale of the hub impacts represented by the number of pupils at primary school hubs, which serves as a proxy for the breadth of effects. The logic demonstrates that, unlike connecting a residential/business property where only a few individuals benefit, connecting a primary school meets the needs of hundreds of pupils through one investment.  

This provides a starting point for estimating the gross impact, i.e. 145 pupils benefit per connected primary school. The RGC evaluation also used the financials of schools to monetise the use of IT in the schools, finding an annual expenditure of £75k on software, broadband charges and other non-capital IT products, alongside capital investments in equipment. 

We think this provides a basis for monetising impacts. Our first approach to monetise impacts is to use difference-in-difference analysis to estimate the change in expenditure due to the broadband connection. This would analyse the change in expenditure after the Hub connection (the first difference over time) and compares this to a counterfactual (the second difference). We will also establish a value to for the shift to IT expenditure in schools’ budgets. This indicates a decision to reallocate resources from other priorities to IT following the broadband connection, providing a so-called revealed preference for EdTech or administrative technologies. This can be used as measure of the value placed on improved technology.   

Whether the approach can be extended to GP Surgeries and libraries/community centres is a key question. As with school hubs, qualitative research with the Hubs can be used to establish the likely resourcing effects after a broadband connection, identifying technologies adopted and their likelihood of appearing in secondary data. This would provide evidence about the logic model for different types of hubs.  

Data will be collected using two methods.  Firstly, it will use accounting data where IT spending data is available. This was the approach used for schools, and there may be equivalent data for other types of Hubs. Discussions with Department for Health and Social Care may be a first step for the health centres and GP surgeries. For libraries early discussions indicate libraries are not individual reporting entities, but part of library services within the local authority. Their accounts are not available. Secondly, it will conduct a baseline and follow up survey. This would include questions on expenditure on IT (if not available through accounts) as well as questions about technology adoption.  We would also consider the potential to either survey a comparator set of Hubs or identify an alternative source of evidence about the GP surgeries or libraries that have not been connected through Hubs.  

A strength of both approaches is that by focusing on resource switching, the approaches align with the efficiency focus on the theory of change. The “more for less” analysis also lends itself to monetisation because it uses metrics that are already expressed as values or can be easily converted into prices.  

The approaches also avoid assessing the outcomes of the public service provided at the Hubs. While it would be desirable to measure these outcomes (e.g. educational attainment), the impact pathway is too complex and subject to too many other influences to be feasible. These approaches are more likely to show a clear link between the broadband investment and measurable outcomes that can be analysed quantitatively. 

4.6 Time savings through tele-consultations

An additional strand of analysis would be focused only on GP surgeries. This approach could explore whether the upgraded connection has affected GP practices’ ability to conduct consultations online, which could generate time/cost savings for the practice and for patients. Then, the approach would be to monetise impacts from a cost savings perspective. The distinct impact pathway would look at the activities enabled by the connection (and any associated expenditure on IT), and cost the activities, identify any changes and look at differences. This would evaluate whether the upgraded broadband connection has improved GP practices’ ability to conduct online consultations, thereby generating time and cost savings for both the practices and their patients. 

This would involve the following steps: 

1. Using difference in difference analysis, compare the changes in the number of online consultations before and after the broadband upgrade, between GP practices with upgraded connections and similar practices without an upgrade.  This would establish the causal effect of the investment on the number of remote consultations undertaken.  

2. Identify the time/cost savings associated with online consultations relative to face-to-face consultations. By combining with the DiD analysis we could then estimate the value of time saved because of the Hub investment.   

3. Estimate the time saved by patients using online consultations and assign a monetary value to this using benchmark on the value of time (e.g. those that are used for appraisals of transport investments). This would allow us to estimate the value of time saved for patients as a result of the Hub investment.  

This could involve the use of primary and secondary data from a range of sources: 

Time series data on online consultations for individual GP practices can be accessed from the NHS Digital Data Link: General Practice Appointment Data Practice Level Breakdown Summary May 2024. This dataset provides detailed information on the number of online consultations. 

We would conduct a literature review of existing studies that quantify the cost/time savings associated with online consultations. These studies should provide benchmarks and methodologies for monetising the benefits. 

This could be complemented by qualitative research with GP practices to understand the operational changes and time savings resulting from increased online consultations. 

We believe there is value in exploring the impact of upgraded broadband connections on GP practices’ ability to conduct online consultations. This is a distinct and direct impact pathway resulting from investment in broadband which should provide valuable insights.  This approach aligns with the overall goal of evaluating the comprehensive benefits of upgraded broadband connections and supports informed decision-making for future investments. 

Hub impacts beyond the Hub

The empirical work for RGC considered the uptake and availability of fast broadband in areas near Hubs. This involved a counterfactual approach. Initially the expected effects were not observed, likely because more time is needed for these effects to become evident. Future efforts to monetise any impacts will build on methods used in other studies.  

Logic model for indirect impacts

The theory of change developed for the Hubs intervention’s indirect effects is, again, outlined at a high level. A useful framework for understanding these wider effects is the model of spatial dimensions, which is described in other evaluation documents. Essentially, the Hub serves as a central point from which broadband connections gradually extend. Suppliers would be able to use the Hub connection to connect nearby residential and business properties. That would then bring a further set of potential connections. 

The model recognises that additional public support may be required. In rural areas, the distances to the periphery of a town or village and the relatively low density of properties increases costs beyond what suppliers can cover commercially. Here, vouchers and voucher projects can help bridge these cost gaps. 

Approaches to monetise

Monetising indirect impacts involves focusing on the connections to properties which are attributable to the Hub. The RGC evaluation provided methods for quantifying these additional connections, consistent with the model of spatial dimension. There is detailed data at low geographical level about broadband availability and performance. Analysis of this data looked at changes near Hubs with those in equally remote areas without supported public buildings. This approach can identify additional effects in the areas around Hubs which are not seen in the counterfactual. 

If there are connections attributable to the Hubs. they can be monetised using the result/methods of other studies, such as the Superfast Broadband evaluation. This includes:  

• productivity effect of businesses, varying by sector 

• labour market effects (mainly reduction in unemployment) 

• wellbeing effects (using house price change as a proxy indicator) 

There is evidence that applying these methods would be appropriate. The RGC evaluation conducted surveys of businesses and residences near to Hubs and found that the benefits experienced were similar to those in other connected areas (Hatch, 2022, Vouchers Evaluation). This means that the valuation process used in other studies is applicable here.  

There are, however, some methodological challenges. A key one is to avoid double counting the benefits if the wider connectivity was enabled through the other government support (such as vouchers). This can be reduced by controlling for this factor in the estimation of additional effects. 

Data requirements  

The RGC evaluation used the Connected Nations data, which tracks broadband availability and performance in small areas over many years. This was linked to voucher beneficiaries to avoid double counting. Other geographical datasets are also integrated, such as the location of schools, to provide a dataset suitable for statistical matching and small area difference-in-difference analysis. 

In addition, the plan would be to survey residences and businesses near to GigaHubs in late 2024. This could validate the results of the RGC surveys, which confirmed that connections in rural areas had similar effects to those made in other properties, with respondents confirming similar changes in uses, satisfaction and overall wellbeing. Alternatively, these questions may be dropped and the RGC results taken as an assumption for the rest of the evaluation. 

The survey may also be extended to new areas. It may test whether the way respondents heard about the potential to upgrade their connectivity differs from the RGC. Early interview findings suggest that the GigaHub investments have been accompanied by greater local, community level participation/consultation and more supplier marketing than RGC Hubs. The survey could confirm this, and what the drivers were. This may then validate any attribution of the indirect effect and assist the monetisation effort. 

4.7 Approach for the process review of GigaHubs

For GigaHubs, a process review is planned to be conducted alongside the delivery of the programme. This process review focuses on the GigaHub part of Project Gigabit, involving a set of projects commissioned by BDUK. The projects connect an agreed list of sites, which then would deliver public services using the faster broadband connection. BDUK oversaw the development of the individual GigaHub projects, supporting each project through four stages: initiation, scoping, procurement and delivery.  

The following diagram presents the stages, describes each including how long they were expected to take and then indicates the methods to be used to review processes. The stages were articulated in the toolkit provided to potential project leaders as they developed their initiatives and the stages are used to structure the process review. 

The first stage in the process focuses on initiating the project and building collaboration, with a lead authority being identified and partners drawn into the project. Initiation built on successful past collaborations with DfE and NHS Scotland. This includes establishing the lead by defining the lead organisation and project champion and developing a partnership through understanding the needs of potential partners. This stage is expected to take 8 – 12 weeks to complete. Evaluation of this will be achieved through analysis of management information including organisation level modelling and interviews about building collaborations.

The second stage of the GigaHubs process develops the project scope. This identifies the schools, GP surgeries, libraries and other public buildings needing a subsidised connection.  This includes building a site list by identifying potential hubs that meet the criteria and developing a procurement strategy by reviewing procurement and funding options. This stage is expected to take 4 – 6 weeks to complete. Activities that support the evaluation of this include monitoring data about project scoping, interviews about site selection, and the development of projects.

The focus of the third set of processes is execution and commitment. Procurement took about the expected two months for some projects. This includes implementing a procurement route such as issuing an Invitation to Tender (ITT) in preferred market route, and the grant offer award, which depends upon assurances and completion of submission to receive approval. This stage is expected to take 8 weeks to complete. Activities that support the evaluation of this include interviews about procurement including different routes to market and modelling the effects of delays.

Finally, projects have processes associated with the build. To date, only LA- led projects have moved to delivery, and the operational processes involve the appropriate teams being put into place and a good relationship between the suppliers and project leads. There has been a good move from committing to GigaHubs to the build phase. This includes managing the project delivery through the lead body responsible for the management of the supplier, and reporting delivery by completing project management information templates. This stage can take 1 – 2 years to complete. Evaluation of this will take place through interviews about delivery, covering supplier management, covering topics such as supplier management, further connectivity, and interaction with other digital policies.

The process review’s structure mirrors the stages of the GigaHub programme. The figure indicates these stages, stating what each aimed to achieve, and how long each stage was expected to take. The final row in the figure indicates the evidence gathered for this review to assess processes. The review is based on analysis of GigaHub monitoring data and interviews with individuals involved in developing the Hub intervention, leading the projects that are delivering connections, suppliers that are connecting Hubs and those delivering services at the Hubs.  

The review focuses on three projects led by individual local authorities (LAs) for their own areas. There are then three projects connecting Hubs in multiple LAs. A first led by the Department for Education (DfE) focuses on primary schools. The Midlands GigaHub Project, hereafter referred to as the Midlands Project, aimed to connect schools, libraries and other buildings in ten LAs in the central region of England. This project progressed to procurement but is not progressing further as a supplier was not contracted. A final project run by NHS Scotland has five locations across Scotland. 

Methodology

The report collects primary and secondary data to understand the different stages of the programme and evaluate whether each stage achieved its objectives. There is also a focus on understanding the overall process, the wider context in which the Hubs are being connected and identifying the activities which improve efficiency and those which are lacking. Evidence gathering is structured to identify different aspects of the process review: 

Review official documents to understand all stages. The study reviews documents on Project Gigabit to understand the different stages of connecting a Hub, the goals of each stage and the interlinkages between them. This review includes the government’s guide on Project Gigabit and related reports, BDUK’s theory of change (ToC), and policy and strategy reports of four LAs which are a part of the programme. All documents were publicly available on government websites, including that of central government departments and LAs.  

Analysing the management information data. BDUK maintains a record of all Hubs including location, connection of status and local delivery body. The Hub intervention data related to the 2,554 Hubs supported by mid-2024. The management information data was shared with Belmana at seven different time periods: July 2023, September 2023, November 2023, December 2023, January 2024, March 2024, and July 2024. This data has also been linked to other public datasets such as DfE’s data on English schools and Ofcom’s data on broadband availability and use in the UK (called the Connected Nations database) to further analyse the Hub connections. This quantitative evidence includes some modelling of the potential sites in areas that may be candidates for a Hub project. In particular, a list of all English primary schools, GP surgeries and libraries – including the Hubs – is linked to organisations that might lead a project to explore how many sites potentially could be part of a project if led by the organisation.  

Conduct in-depth interviews. Qualitative evidence is drawn from interviews conducted at four case study sites: Dorset, Leicestershire, South Oxfordshire and the Midlands region (Nottinghamshire and Shropshire). Case studies in each of the study sites typically began with an interview with the local delivery lead (LDL) or local delivery manager (LDM) at BDUK. This first set of interviews explored the aims of the projects, the way projects have been scoped, selecting sites and then how connections were procured. BDUK helped involve LAs leading the projects who then further enabled contact with Hubs and suppliers. The interviews were conducted between September 2023 and July 2024. Typically, the conversations were between 30-45 mins long and were mostly conducted online.

4.8 Evidence to answer research questions

The process review aims to assess the efficiency of the GigaHubs in achieving its objectives and identify opportunities to further improve it. It will cover the process aspects of all stages from project initiation to delivery, evaluating the evidence about process effectiveness and identifying the lessons learnt at each stage. This process review specifically seeks to answer or contribute to the understanding of these questions:

• What is the process of connecting a GigaHub to fast broadband?
• How effective and efficient has the GigaHubs delivery been?
• What can we learn to improve future policy designs and implementation?

The review looks at the individual stages answering these questions in terms of each of the stages. A mix of quantitative, qualitative and comparative evidence is collected to see whether the stages individually facilitate benefits and impact and look across the stages to pick out how processes efficiently enable this or where this did not take place. The next sections develop the research questions in terms of the four stages of the the image shown above.

Initiating projects to connect public buildings

A first stage of the GigaHubs process identifies the lead organisation and the partnerships for a GigaHub project. At the time of the process review, the number, scale and scope of the initiated projects will be considered. Key dimensions of this would be whether there are sufficient projects for the delivery expected, looking at the type of organisation that leads a project.

The aspects of the process to considered is how the GigaHubs were marketed and to which organisations. Based on previous similar projects, there would be a mix of project leads, some led by LAs are focusing on a county; others may have a larger area of focus and be led by government departments. The review would consider the target organisations for leading projects and whether this was sufficient for the desired number of Hub connections. There will also be evidence gathered about how BDUK reached out to potential project leads, and worked well and where there may have been improvements.

Crucial to the initiation is the extent to which organisations collaborate and the avenues by which this takes place. The focus for interviews and other evidence gathering would be the objective setting for the projects, articulating the benefits of Hub investments. These would be expected to centre on improving the services provided at the Hubs, but objective setting would also articulate wider community benefits and the development of connectivity in areas distant from the gigabit-capable infrastructure. This strand of project scoping may be linked to the experience of teams with earlier funding of the Hub intervention and the links across and between organisations leading projects to the service provided at Hubs.

The GigaHub processes envisages that projects connecting public buildings might attract funding beyond that secured from BDUK. Evidence about the scoping phase would cover whether and to what extent the GigaHub Projects, as well as delivering GigaHubs, provided a route to connect public buildings more generally, by marshalling in other local or national funding, leveraging further benefits from the project.

Scoping projects

Scoping by project leaders involved compiling lists of sites and then information to determine whether each building met GigaHub criteria. BDUK has collected data to monitor the progress connecting sites including the sites that may initially be viewed in scope but subsequently be descoping. The review will use this monitoring data to the way projects are scoped. the scoping period with projects maturing the list of sites to be connected. Data should indicate the scale of the overall projects and the dynamics of site selection over time.  Crucial would be the interviews with project leaders and these views would then be corroborated/broadened with other stakeholders exploring the actions taken to enhance the site lists.

Overall, the testing of the process would look at the criteria for Hub selection. A focus for projects is the public buildings which were unlikely to be connected commercially. The data relevant to such criteria, such as BDUK data focusing on access to gigabit infrastructure at site level, will be an input into this stage in the process. Views will be sought about its use and alternative evidence that projects could use to establish the connectivity needs of individual sites. Discussions with stakeholders may also outline other criteria such as whether a site would quickly take up fast broadband services after a connection.

A crucial aspect of this stage of the process is the alternative routes to connect sites. Evidence will be gathered about the way site scoping interacted across projects (e.g. a national project might interact with a local one to determine which project include a site). The process review will explore the fact that there are multiple projects and scoping had to consider overlaps across the projects.

Procuring Hubs

After site selection, project leads turn to procuring the connections within the scope of the GigaHub projects. In setting out the processes, BDUK did not advocate a particular route to market. The initial guidance, the “toolkit” for projects, envisaged those leading a project choosing how connections would be procured. There is a recognition that a procurement route may not always exist, and BDUK highlighted an option that could be used that was already in place.

The process review will look at which routes are selected. A consideration would be whether the suite of dynamic purchasing systems for Gigabit-capable Connectivity (RM6095) is used. The strengths and weaknesses of alternatives may lead to different routes for different types of projects.

Questions will then be whether the routes are efficient and could be operated smoothly to procure connections to public buildings. A key consideration for a procurement will be whether – as the project procures connections to public buildings – there is adequate information to allow the bidding suppliers to costs works, whether the structure of the procurement can take steps to assure value for money and the supplier perspective on whether the GigaHub contracts would be attractive. Interviews would also then need to context the wider procurement landscape, especially any interaction with the larger GIS procurements operating in parallel.

Delivering the GigaHubs

Delivering connectivity to the Hubs will involve project management processes, especially as the project leads interact with suppliers. Interviews with key stakeholders, suppliers and project leaders will cover how each project has been managed. It would also look at the relationship between the projects and the Hub sites.

The aspects of delivery that will be tested in this evidence gathering will be both contextual and look at the delivery management actions. By way of context, the past track record of delivery and collaboration in comparable projects, the nature of the supplier arrangements, and whether the understanding of the various participants about what works and likely delivery problems will be considered. In the delivery actions, as well as the routine management actions tracking delivery and resolving delivery issues, the extent to which key dependencies are managed will be explored. The relationship between delivering infrastructure and local authority planning systems has been highlighted as an issue in past evaluations, and this process review can explore steps that worked to ameliorate these issues or consider improved ways of working.

The capabilities, dependencies and risks in the Hubs projects will be explored, focusing more on the management teams looking at how these are managed. There will be a consideration of the skills/experience needed in project teams, the communication channels employed, the interaction with the hub sites, and the shaping of the Hub projects within the wider gigabit-capable network rollouts being delivered. These aspects will then be tested with the suppliers, to validate and contrast this aspect of delivery

A crucial aspect of delivery will be the processes to capitalise on enhanced connectivity at public buildings. Some of the benefits derived from the Hub connection would be delivered only if complementary digital technology investments are made. Stakeholders interviewed will include those involved in such parallel digital policies, such as LA digital inclusion leads, or the policy leads on educational digital technology. How the projects interacted with the parallel investments will be collected through qualitative evidence.

The main quantitative evidence about delivery is the extent to which connections have been delivered and whether this would be to time. The GigaHubs monitoring data would track the connections and establish whether these are on time. A further quantitative dimension would be the resources involved in securing the Hub connections. The costs of connections will generally be set out in the contracting and – perhaps more meaningful – would be to establish whether and where procurement of sites did not occur, to explore if the decision not to deliver site connections reflect a process issue or the connections not being value for money.

Assessing the GigaHubs process

The staged approach to this process review offers an opportunity then to assess processes in the round as well as in terms of individual. A key consideration will be whether the GigaHubs projects have delivered the level of connectivity that was expected, and this was to time and budget. This builds on the delivery stage findings. Where there were delivery issues or successes, the review would then look back to the earlier stages to explore whether the performance levels could be explained and the lessons learned at the different stages.

At the initiation stage, the main drivers for project success would be project leads and collaborators and having leadership to progress a project to completion. At the scoping stage, the identification of a list of public buildings would depend on good information about connectivity issues and links with public sector delivery bodies that operate in the public buildings. The testing of whether the knowledge to provide scope to projects quickly and accurately would be crucial to review.

This would also factor into whether an appropriate procurement strategy, that can be put to the market, is developed. Progressing that procurement strategy in the execution stages depends on the routes to finding suitable suppliers being suitable for projects and contracting in a manner that assures delivery. The review can gauge strengths and weaknesses both in GigaHub processes and then testing these in the context of alternative routes.

Findings will be presented in a report, mixing the qualitative and quantitative findings, and structured in a manner consistent with the staged process for GigaHub projects. Conclusions will focus on the GigaHubs processes but then also read across to the other related programmes to improve connectivity and the use of digital public services in remote areas. 

5. Implementation

This section sets out how the rest of the Hubs evaluation will be implemented. It draws on the information provided in the preceding sections of the report. The section includes the timelines for each evaluation stage and the key risks for the evaluation. 

5.1 Updated evaluation timetable

The diagram below shows the updated evaluation timescale for this study. The timescales of key milestones are as follows, aligned with reports to be completed. The publication of reports would then follow after.

Milestone 1: Report BDUK Gigabit Hub Evaluation: Process Review. Examining how effectively the GigaHub programme – a set of projects that delivered Hubs starting in 2021 – has been implemented and the key learnings from the delivery of the programme (December 2024). 

Milestone 2: Initial Impact Evaluation: GigaHubs and RGC. Reviews the impacts of the GigaHubs by the end of 2024, after the delivery concluded, and updates findings for RGC Hubs (September 2025). 

Milestone 3: Interim Impact Evaluation: GigaHubs and RGC. Updates the impacts of the GigaHubs and RGC Hubs by the end of 2025 (August 2026). 

Milestone 4: Final Impact Evaluation: GigaHubs and RGC. Reviews the impacts of the Hub intervention across projects by the end of 2026, three years after the delivery concluded (April 2027). 

The first milestone’s focus is the processes used to deliver the GigaHubs. The further three milestones then report impacts for both RGC Hubs and GigaHubs. These studies will mature the evidence both in terms of the years of impacts assessed and the range of impacts covered. The figure indicates how strands of research method will contribute to the milestones 2-4. The focus for the milestone 2 will be the first analysis of GigaHubs, looking at evidence about one-year impacts at these Hubs, with evidence primarily from in-depth interviews and the area level data analysis which will cover all Hubs. At milestone 3, that quantitative analysis will be updated for a final time, to capture three-year effects. There will also be a more complete picture of the library Hubs effects. 

Milestone 4 will be the final evaluation drawing together the full range of evidence, including the overall assessment of the value for money of the Hubs intervention.  

Table: updated evaluation timetable

5.2 Risk Register

This section presents a risk register that will be monitored during the project. The risks will be mitigated using steps presented in the table. 

Table: risk register

Risk	Likelihood	Impact	Mitigation
Issues identifying a suitable counterfactual	Low	High	First stage analysis has developed methods Further analysis structured to build on the analysis in place
Evaluation findings do not stand up to external scrutiny	Low	High	Routine, standardised methods with established codes and data BDUK steering group input at key stages
Low response levels to survey	Medium	Medium	Monitor response rates and follow up with telephone contact Over-sample on surveys Design surveys to be response friendly, pilot / test for this
In-depth interview contacts unavailable or unresponsive	Medium	Medium	Work with BDUK, project leads and others to ensure first contact as warm as possible  Widen the paths to recruit individuals by following up residential and business surveys and Hub surveys  Design research frameworks and time interviews appropriately
Loss of access to secondary data	Low	Low	Data sharing arrangement in place Routine dialogue about data and key Reshape analysis if risk materialises
Delays in programme delivery	Medium	High	Monitor delivery and routinely discuss plans Replan the project as the delivery pace becomes clearer
Resources unavailable at BDUK	Low	High	Relatively long project so that gaps in staffing can be managed Research team at BDUK is two-deep allowing re-allocation of tasks
Resourcing issues at supplier	Medium	Medium	Relatively long project so that gaps in staffing can be managed Consortium reach back and working in a number of related studies means team members have back up working in similar projects

At this stage, replanning the study, some risks have materialised with regard to ‘delays in the programme’ and the replan reflects this. Also, the in-depth interviewing has been reshaped. Original plans were for the case studies to be centred on and in local authorities. As the area level operation of the Hubs varies across projects – some projects have an area focus but many are national – the recruitment of the residential and business interviewees has been through surveys rather than project interaction. In addition, some Hub interviews will be secured through Hub surveys. 

1. Belmana, with Hatch and Winning Moves, BDUK Rural Gigabit Connectivity Hubs Evaluation, December 2023. Available from: BDUK Rural Gigabit Connectivity Hubs Evaluation) Accessed 6 August 2025 ↩