Measured productivity fell dramatically during 2020 as the pandemic changed how public services operated and led to a reduced quantity of ‘normal’ activity. This was most clear in health care, where almost all non-emergency procedures were put on hold and the activity of the service focused on dealing with the pandemic. However, it also happened in other services (for example, fewer court cases were heard). This led to a very large fall in measured productivity. In 2020, overall public service productivity fell by over 15%. It only recovered partially in 2021, remaining 9% below 2019 levels and 5% below 1997 levels.

How public service productivity has recovered since then is still uncertain, in part because the ONS’s official statistics, incorporating a full quality adjustment, are only published with a three-year lag. Other research has shed some light on the state of public services more recently and presents a gloomy picture. Relative to 2019, the government has put a lot more money into public services, increasing inputs. However, in some services – healthcare in particular – there has not been a commensurate increase in activity. Between 2019 and 2023, staffing levels in the NHS increased by almost 20%, but key measures of activity remained broadly flat. ²⁷ www.ifs.org.uk/articles/there-really-nhs-productivity-crisis  Performance Tracker identifies a similar, though not as stark trend, in other services like the police.

While the fully comprehensive ONS productivity measures are not yet available post-2021, the ONS has provided some less comprehensive but more up-to-date estimates. Using quarterly data on service activity (not adjusting for quality), the ONS estimates that productivity was around 6% lower in 2022 than 2019 (and around 2% lower than 1997). They have also presented an alternative ‘dynamic regression’ estimate, which attempts to capture how quality might have changed based on past annual data, and implies productivity in 2022 was similar to 2019. However, this method, which was produced before the 2021 full data was available, was overly optimistic about the recovery of productivity in that year, ²⁸ The dynamic regression suggested productivity in 2021 would be 2% higher than 1997 levels, with a lower confidence interval of 0.4% below. The final value was over 5% below 1997 levels.  and it would be surprising if quality has improved as much as the ONS’ dynamic estimate implies. As a result, until full annual quality-adjusted data is available, the quarterly data is the best guide and suggests productivity remained much lower in 2022 than it was before the pandemic.