Productivity green shoots boosted by financial services

After a long period of stagnation, between Q3 2024 to Q3 2025 UK labour productivity grew by 1%. This modest growth is welcome. But the improvement might be larger as this figure is debated due to the fall in responses to the Labour Force Survey (LFS); the denominator of this labour productivity measure. The Office of National Statistics (ONS) now provides an additional productivity metric using the number of employees derived from HMRC PAYE data. This measure suggests labour productivity grew at 3% instead.

There are, however, uncertainties around both figures. While the HMRC PAYE data is more accurate for the number of employees, it doesn’t provide data on self-employed or hours worked – both of which continue to be collected via the LFS.

The challenge for productivity analysts is that only the LFS dataset enables productivity to be assessed from a bottom-up perspective which provides an insight into what is happening across sectors. And the LFS sectoral data reveals some important shifts that have taken place since 2019.

Between Q3 2019 and Q3 2024 private sector labour productivity shrank by -1.8%, but it grew between Q3 2024 and Q3 2025 by 0.2% (Table 1). Moreover, just over half of the private sector contributed positively to productivity growth over the last year, whereas only just over a third contributed positively between 2019 – 2024.

While this shift is positive, it is important to note that most of the 1% headline growth in the most recent period was due to the public sector “Between Effect” caused by an increase in its labour share. However, the public sector will most likely begin to generate a negative “Between Effect” as the civil service is expected to cut headcount by 8%, with people likely moving to lower value-added services roles. In any case, the policy focus must remain on driving private sector productivity growth.

Productivity grows when firms are able to increase value added per hour by creating a competitive advantage in the way labour and capital of all kinds are deployed, and by increasing the labour share in higher value-added activities. While it is positive that the majority of private sector industries are now contributing to growth (in green), it remains a concern that both Mining & Quarrying and Manufacturing (both of which are high value-added) continue to contribute negatively (in red). The data also suggests that government attempts to revitalise Manufacturing through its industrial strategy have so far had a limited impact.

Table 1: Sectoral productivity disaggregation Q3 2024 – Q3 2025^[1]

The high value Financial Services sector contributed most to recent productivity growth due to its increase in labour share. Between Q3 2024 and Q3 2025 exports of Financial Services grew by 7% with the Insurance sector growing at 8% indicating a strong demand for these services (Chart 1). Conversely, goods exports decreased by 5% over this period.

Chart 1: UK Exports of Insurance and Financial (banking) services 2016-2025 (current prices)

Source: ONS

By far the largest export destination for traditional insurance services is the US – accounting for 42% of exports followed by the EU at just 16% and Canada & Australia at 14%. In 2016 when the UK was part of the EU, exports to Europe accounted for just 15% of exports, highlighting the fact that there was no single market in financial services for UK firms to exploit. Each member state typically has its own regulator, rules and product requirements.

The digital transformation of the global economy is, however, unlocking new opportunities for the UK insurance sector. For example, the market to insure cyber risk is estimated to be worth $50bn by 2030 which is largely dominated by US firms. The UK has an underdeveloped cyber insurance market despite the fact that the UK faces more cyberattacks than any European nation which is a significant opportunity.

Further opportunities for the UK insurance sector are being created by firms seeking insurance for their AI products and services including autonomous vehicles, robotics, AI‑Driven Healthcare and Diagnostics ad well as for Generative AI. But to achieve a growing market share will require the sector to clearly understand the risks involved and levy an appropriate premium that is both attractive for firms while managing future liabilities.

The challenge for the insurance sector though is that there isn’t a history of loss data upon which to develop new models, and hence the industry will have to develop new approaches to pricing this risk. This includes managing systemic risk in the event that a widely used AI model such as a large language model fails, which in turn forces thousands of companies to fail and potentially the insurance company too.

Each frontier sector that is contained within the IS-8 has the potential to drive growth. Indeed, the government’s Financial Services Growth and Competitiveness strategy references the need “to make the UK the location of choice for insurance and reinsurance” with a focus on regulation. But if the UK insurance sector is to scale up and become a global leader in AI and cyber insurance, it is the insurance sector itself that will need to draft a plan to grow not the government.

Although the green shoots of a productivity recovery are beginning to emerge, it is too early to tell whether UK firms have indeed found new forms of competitive advantage to compete in global markets. The UK insurance sector certainly has the deep expertise and skill to scale up in new markets such as cyber security risk and AI should the sector decide to come together and formulate a coherent growth plan.

^[1] The sectoral disaggregation uses GEAD (Generalized Exactly Additive Decomposition) based on Tang & Wang (2004). The ‘within’ effect is productivity growth in activities within the sector whereas the ‘between’ effect measures the change in relative size of sectors taking into account the reallocation of labour between sectors and changes in real output prices. e = estimated as ONS does not publish these values.

The views and opinions expressed in this post are those of the author(s) and not necessarily those of the Bennett Institute for Public Policy.