Introduction
A few years ago, researchers at Sheffield Hallam reviewed the energy strategies produced by Local Enterprise Partnerships and found that hydrogen featured in roughly half of them. Not half of the places with a refinery, a salt cavern or a heavy industrial cluster, half of the total number of strategies produced.
The observation is not offered to score a point. It illustrates a pattern that anyone working in sub-national economic development will recognise, and which has, if anything, intensified since then with the publication of the Industrial Strategy. Read five sector strategies from five different authorities, cover the place name and see whether you can tell which is which.
It is an easy trap but needs avoiding
The drivers for identifying the same sectors that everybody else has come up with are, in many ways, rational. The Modern Industrial Strategy identifies eight priority sectors — the IS-8 — and directs support towards them. Local Growth Plans are intended to be the guiding star for mayoral strategic authorities and their partners, and are expected to align with the national picture. As a consequence, many Combined Authority Sector Development Plans map closely onto the IS-8, and the documents say so openly. Where funding, ministerial attention and departmental engagement are organised around eight headings, it takes a certain nerve to write a strategy that sits outside them.
Add to the national noise internal dynamics. Consultation with local businesses tends to produce a longer list rather than a shorter one. Members are understandably reluctant to see a sector removed from a published document, since doing so reads locally as a policy to ignore it. And no economic development director wants to be the one who took clean energy out of the strategy the year before a major grid investment lands.
The result is a familiar compromise: a strategy that names six or seven sectors, describes each in broadly positive terms, and commits to a proportionate share of a limited resource across all of them.
What it costs
The approach, however, is almost impossible to succeed with but, importantly, creates three damaging effects.
The first is dilution. A small team cannot service seven sector propositions. In practice, one or two get real attention, and the rest get a webpage.
The second is credibility with investors. Site selectors and corporate real estate advisers read a great many of these documents. They are looking for specific evidence, occupier types, available power, labour supply at particular occupational levels, a named supply chain and they are quick to discount generalities. A proposition that could belong anywhere tends to be treated as belonging nowhere.
The third, and increasingly the most material, is credibility with government. As place-based approaches to appraisal and funding develop, departments are working directly with places on coordinated packages of investment grounded in local economic analysis. That is a considerable opportunity for authorities with a clear, evidenced story about what they are for. It is an uncomfortable environment for those without one.
What actually counts as evidence
There is no single indicator that provides irrefutable evidence on its own. What works is a layered picture, where each layer can be misleading, but multiple triangulated data points start to show a compelling story.
Location quotients. Location quotients remain the sensible starting point, but they are routinely over-read. Three cautions are worth stating plainly. First, granularity: an LQ calculated at broad sector level tells you very little, and the interesting findings almost always appear at three-digit SIC codes or lower – but at this level the survey-based data becomes less and less reliable as the sector definition is refined. Second, base size: a location quotient of 1.6 in an activity employing 350 people is a statistical observation, not a cluster, and is frequently the shadow of a single employer. Third, and most importantly, concentration is not competitiveness. An area can be highly specialised in an activity that is declining nationally and losing share locally. Pairing LQ with an analysis which separates national growth, industry mix and local competitive effects usually tells a far more honest story, albeit often a less flattering one.
Occupational data. A widespread issue is that industrial classification tells you what firms say they do (and they are frequently wrong). Occupational data tells you what people in the area can actually do, and these activities frequently cross sector boundaries in ways that reveal real transferable strengths. Online job posting data, for all its known biases, adds a timeliness that official statistics cannot.
Business knowledge. In the end, to understand supply chain depth, one needs to be able to identify and understand the firms operating in a market space. This necessarily goes beyond the one or two anchor employers, but the thirty or forty businesses in the second and third tiers, what they make, who they sell to, and whether they sell to each other. Depth is what distinguishes a cluster from an agglomeration of unrelated firms that happen to share a postcode. Company-level data, procurement records and a modest programme of business interviews will establish this far more reliably than any secondary dataset.
Innovation Mapping. In today’s economy, productivity is often used as shorthand for value-add and knowledge-based assets. Patent mapping can help mine the competitiveness of a cluster, but it is also easy to misuse. Propensity to patent varies enormously between sectors, so raw counts flatter engineering and pharmaceuticals and badly understate software and services. Assignee addresses often record a corporate headquarters rather than the laboratory or development centre where the work was done. Looking at technology classes rather than totals, at inventor location where available, and at co-assignment patterns showing who collaborates with whom (as R&D grant recipient data also reveals), patent data can reveal specialisms that employment data misses entirely. Read alongside wider measures and an understanding of wider innovation assets, it becomes a reasonable proxy for whether an area is generating capability or merely hosting it.
Supply-side mapping. Grid capacity, consented land, port access, test and trial facilities, and regulatory arrangements are among the supply-side factors that can indicate early potential for growth.
Investment data. Undeniable additional evidence comes from what firms have actually done. Where has inward investment landed, where has venture capital gone, which local firms have chosen to expand on site rather than elsewhere. Revealed preference is worth a great deal more than stated ambition.
Three questions before launching a sector-based strategy
- Could a neighbouring authority publish this document by changing the place name?
- If the single largest asset closed, would the strategy still hold?
- Can we name the businesses that a sector strategy will impact and how are they involved?
Specific actions may be better targeted below the sector level
The most defensible strategies rarely claim a whole sector, they claim a position within one.
It is difficult to sustain a claim to be a life sciences region. It is entirely possible to demonstrate a concentration in clinical trial support, in medical device manufacturing serving a particular clinical area, or in the specialist cold chain logistics that surrounds both. Similarly, few places can credibly claim hydrogen production at scale; rather, more can evidence a genuine supply chain position in valves, pipework, storage vessels, or the electrical engineering that hydrogen infrastructure requires wherever it is ultimately built.
This reframing tends to be more persuasive to government, more useful to investors, and considerably easier to deliver against. It also has the advantage of being defensible in a challenge because it rests on things that can be counted.
A sub-sector can be used to carefully identify companies with whom proactive targeting work can be undertaken and the proposition is already defined.
Saying what you are not
The final discipline is the hardest one politically. A strategy gains credibility from what it declines to claim.
That does not require ‘abandoning’ sectors, but there is a real difference between a competitive advantage the authority intends to invest in building, a foundational sector that employs a great many residents and needs support of an entirely different kind, and an emerging opportunity that is being monitored rather than backed with significant resources.
Setting those out as distinct categories, with corresponding interventions, allows a strategy to clarify priorities without giving the impression that any sector does not matter. Such an approach gives relevant groups a place in the strategy while still enabling hard prioritisation choices.
Conclusion
The evidence base for serious cluster analysis is more accessible than ever, and the analytical techniques involved are not exotic. What is scarce is the willingness to let the evidence narrow the answer.
A sector strategy that names three things and can prove them will outperform one that names seven and simply asserts them. In the end, poorly defined strategies simply fail as they cannot meaningfully engage with the local businesses and investors whose behaviour will ultimately determine whether any of it was true.
Mickledore prepares sector and cluster strategies, competitive advantage assessments and sector narrative refreshes for local authorities, combined authorities and their partners. Our work combines granular industrial and occupational analysis, patent and innovation asset mapping, and structured business engagement to establish supply chain depth. If you are refreshing a sector proposition or building the evidence base for a Local Growth Plan, we would be glad to discuss it. You can reach us at nwilcock@regionaldevelopment.co.uk