England’s productivity problem is more tangled than the familiar North-South shorthand suggests. Fresh analysis of official subregional productivity data published by the Office for National Statistics (ONS) shows that some of the widest gaps in output per hour sit between neighboring districts inside the same region, not between opposite ends of the country. The finding challenges a generation of policy thinking built around a simple geographic divide and raises hard questions about where investment should actually go.
What the Latest ONS Data Actually Shows
The ONS publishes annual labour productivity indices for every local authority district in England, measuring both output per hour worked and output per job filled. When analysts apply pattern-detection techniques to that granular dataset, the results scramble the conventional map. Large productivity contrasts show up between neighboring districts, between metropolitan cores and their surrounding rural areas, and between towns that sit just miles apart but operate in different economic orbits.
A companion dataset covering ITL2 and ITL3 subregions, published in June 2025, makes the picture sharper still. It shows that dispersion within regions is substantial. High-performing ITL3 areas turn up in the North, and low-performing ones sit in the South. That pattern is not new, but the scale of the within-region spread has been underappreciated in public debate, where “levelling up” rhetoric still tends to treat the North as a single underperforming block.
The headline numbers also mask important differences between productivity measured per job and per hour. Areas with high levels of part-time work or multiple jobholding can look different depending on which metric is used. Yet on both measures, the core picture is consistent: deep pockets of underperformance sit alongside highly productive neighbors, often separated by little more than an administrative boundary or a stretch of motorway.
Scorecards Expose Winners and Losers Within Regions
The ESRC-funded Productivity Institute has turned the raw ONS numbers into a set of ITL3 scorecards designed to compare places within their parent regions and track change over time. The 2025 scorecards identify which small areas are catching up to their regional peers and which are losing ground. The distinction matters because two districts with similar absolute productivity can be on very different trajectories, one accelerating and the other stalling, and blanket regional policy treats them identically.
These scorecards combine level and growth indicators to classify local economies into broad categories: leaders, improvers, strugglers and decliners. A northern city-region might contain all four in close proximity, with an urban core improving rapidly, a ring of stable but stagnant suburbs, and a struggling former industrial town on the edge. The same pattern can be found in the South West or East Midlands, undermining the idea that geography alone determines fortune.
The Productivity Institute’s separate analysis of Mayoral Combined Authorities adds another dimension. It contrasts productivity level against productivity growth, finding that some combined authorities record high absolute output per hour but near-flat growth, while others sit at middling levels but are expanding faster than their national peers. Even within London, borough-level variation is significant, with major differences in both level and trajectory across the capital. That internal spread within a single city-region is a direct rebuttal to anyone treating London as a uniformly productive powerhouse.
Why the North-South Frame Misleads
The standard critique of English economic geography frames the problem as a prosperous South dragging along a lagging North. That story contains a kernel of truth at the broadest regional level, but it collapses once you zoom in. The ONS statistical bulletin covering 2019 to 2023 explicitly notes that smaller geographies display volatility, and the agency provides both smoothed and unsmoothed series to help analysts separate signal from noise. Even allowing for that volatility, gaps between districts inside the same region remain evident in the ONS series.
ONS methodology constructs subnational productivity as gross value added divided by either jobs or hours, covering ITL1 through ITL3 geographies and local authorities. That approach means the indices capture where economic value is generated, not where workers live, which can create apparent anomalies. A district dominated by a single high-value employer may look spectacularly productive while its neighbour, home to the same workers’ households, looks weak. Understanding that measurement quirk is essential before drawing policy conclusions from the numbers, yet it does not erase the core finding: the gaps are real, and they cut across the North-South line.
Another problem with the simple regional frame is that it encourages zero-sum thinking between broad areas rather than a focus on specific bottlenecks. If policymakers assume the South is already “sorted”, they may overlook coastal towns or ex-industrial communities in the South East that are falling behind. Equally, treating the North as uniformly deprived risks ignoring the assets of high-performing city centres and university-linked innovation districts that could anchor wider growth if connected more effectively to their hinterlands.
AI Skill Demand Follows the Same Fractured Pattern
A peer-reviewed study mapped local demand for AI skills across Great Britain from 2017 to 2022 at small-area and local authority scales. The research found pronounced spatial heterogeneity in AI-related labour demand, meaning that the geography of emerging technology jobs does not simply mirror old industrial boundaries. Pockets of high AI skill demand appear in northern cities, while parts of the rural South show little activity.
That uneven distribution of AI-related work is consistent with the productivity picture. If the next wave of economic growth is partly driven by AI adoption, and if that adoption clusters in specific local economies rather than spreading evenly across regions, then districts already falling behind their neighbours could face a compounding disadvantage. The risk is not that the North will be left behind by the South. It is that a handful of well-connected urban nodes, scattered across both halves of the country, will pull further ahead of everywhere else.
The same study highlights that AI-intensive vacancies tend to co-locate with existing knowledge sectors and strong digital infrastructure. Where those foundations are absent, the pipeline of AI work is thin, regardless of whether the area sits above or below the Humber. This again underlines that the key divide is between connected hubs and disconnected places, not simply between compass points on the map.
What This Means for Economic Policy
The practical consequence is that policies designed around broad regional transfers may miss their targets. A district in the South East that is losing ground relative to its neighbours needs a different intervention than a northern city already catching up. The ITL3 evidence base and the related AI-demand analysis together point towards a more granular approach: identify specific local constraints, then tailor support to unblock them rather than assuming a standard regional package will work everywhere.
One implication is that devolved institutions will need better data and analytical capacity. Metro mayors and combined authorities cannot rely on coarse averages if they are to make informed investment choices. They require tools that can interrogate subregional productivity, sector mix, commuting patterns and technology adoption at the level of individual districts or even neighbourhoods. The emerging ecosystem of scorecards and dashboards is a step in that direction, but it will only change outcomes if it shapes budget allocations and project design.
Another is that central government will have to accept more variation in policy mixes across England. If some low-productivity districts are held back mainly by poor transport links to nearby high-value labour markets, infrastructure spending may deliver the biggest payoff. Elsewhere, the binding constraints might be skills, planning rules or a lack of suitable commercial space for growing firms. Treating all “left behind” places as interchangeable risks wasting scarce public money.
None of this means the North-South story is entirely obsolete. Regional averages still matter for understanding broad imbalances, and inter-regional transfers will remain part of any serious attempt to rebalance the economy. But the new wave of subregional productivity data, combined with fine-grained analysis of AI skill demand, makes clear that the real battle lines run within regions and even within city-regions. Policymakers who ignore that complexity will struggle to move the dial on England’s stubborn productivity gap.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
