A government tender has revealed plans to develop a “skills taxonomy” that could sit at the heart of Department for Education proposals to dictate college courses.

Ministers want a new algorithm developed to map England’s skills shortages, with experts predicting it will play a key role in putting local skills improvement plans (LSIPs) together.

Beginning with a £25,000 research review by Frontier Economics, the taxonomy is intended for the education secretary’s new Skills and Productivity Board to identify and analyse national skills needs and those that “add the most value” to the economy.

The board, composed of seven labour market and skills economists who have been tasked with influencing policy, has met four times since it was formed in November but minutes of the meetings have so far been kept under wraps by the DfE.

The skills taxonomy tender appears to be the first insight into the board’s direction of work.

It states that the board requires a “functional and operational skills ‘taxonomy’ – to be able to name, define and classify skills (for example, knowledge skills, workplace skills, transferable skills)” as well as a way of linking these skills to occupations and, possibly, qualifications.

The board is then expected to use this taxonomy “to identify areas of significant skills shortage”.

Labour market economist and chief executive of the Federation of Awarding Bodies, Tom Bewick, said taxonomies can be useful in terms of “helping to shine a light on trends relating to occupational skills gaps and shortages”, but warned they are “hopeless at predicting employer or student demand in a fast-changing economy”.

Similarly, Stephen Evans, who heads up the Learning and Work Institute, said a skills taxonomy can help commissioners and providers plan provision, but added there has “been so much work on this already and ultimately a taxonomy is no substitute for increased funding for adult skills”.


What is a skills taxonomy?

A skills taxonomy tries to set out what skills are needed in particular roles. It is essentially a skills map for different jobs and occupations. That then is intended to help skills providers design provision and to help people understand what other jobs their skills might be suitable for.

Globally, work has already taken place on skills taxonomies which have typically linked occupations to skills. The O*NET database in the US is the most famous.

The DfE’s tender also points out that innovation agency Nesta and the Economic Statistics Centre of Excellence recently undertook an exercise using “web scraping techniques for online vacancy data to produce their own data-driven taxonomy of skills demand in the UK”.

Part of Frontier Economics’ brief is to bring together the work currently taking place on skills taxonomies in the UK, including techniques such as information from web scraping vacancies and other real-time information.

Labour market analyst firm EMSI does a lot of economic modelling for the FE sector, including scraping job posts to inform colleges what employer demand looks like in their area for them to then translate that back into curriculum planning.

EMSI’s managing director, Andy Durman, said this is one piece of the jigsaw that gets beyond the headline figures that government data can show, building a “more specific insight” into which particular skills occupations require.

He warned, however, that web scraping job postings is “not an exact science” and it is “fraught with a lot of danger” as they “crop up in multiple places”.

Durman urged the DfE to be “very careful to assume that job postings tell us everything about what’s needed”.


How could the skills taxonomy be used?

Bewick told FE Week he believes the algorithm will help form the education secretary’s evidence base to intervene in geographical areas where “they believe FE providers are not sufficiently shaping course provision to meet the aims of LSIPs”.

He said it looks like government wants a new national framework for deciding whether or not local course provision meets real labour market needs, and such a taxonomy “could be what all local areas are mandated to use when putting together LSIPs in future”.

LSIPs are central to the DfE’s FE white paper reforms. They will be piloted this year, and legislation is being drawn up to put employer leadership of the plans on a statutory footing.

Colleges and training providers must then pay “due regard” to the plans by ensuring their courses align to local employers’ needs. Failure to do so will result in intervention from the education secretary.

Bewick warned that if the intention of the DfE’s skills taxonomy is to try and predict how many hairdressers or mechanics a local area needs to train, “it will fail”.

“Moreover, post pandemic, there will be new jobs and skills created which labour market economists will not be able to anticipate or fully capture in their econometric models. In simple terms, this means that any taxonomy tool will already be out of date before it is even used.”

Durman agreed. He said the world moves at pace and things decay. “You can’t afford to pay massive panels to keep tweaking it every few months so they tend to be fairly static.”

Linking skills shortages to which quals are made available has proved controversial in recent months. The Education and Skills Funding Agency sparked outrage when it excluded major sectors including hospitality, tourism and media from its new level 3 adult offer as they were deemed to be a low priority with low wages.

Evans pointed out that any new skills taxonomy would not be ready until LSIPs are already in play, but they could be used to inform them in future.

He also pressed that the government needs to be “much clearer” on how LSIPs fit with Skills Advisory Panels, Local Enterprise Partnerships and local government, otherwise “we risk an alphabet soup of acronyms”.

Frontier Economics’ contract started in April and is due to end in July. The initial findings are expected to form the basis for a larger piece of research in the next financial year.

Your thoughts

Leave a Reply

Your email address will not be published. Required fields are marked *


  1. There are some interesting fundamentals beneath all this, but essentially it’s a rebranding exercise, with a coercive undertone bolted on.

    It’s based on the premise that filling skills gaps will be economically beneficial (as opposed to the ‘education’ model being for personal fulfillment, where economic benefit is a by product).

    It assumes that the employer led system will be more beneficial than a learner led model, which is unproven and likely unprovable! – but because of favouring loans as the funding model, the learner is taking on the financial liability (ie is the customer) and the customer chooses where and what to spend on… right?

    It’s really about shifting financial liability away from the state and onto individuals, whilst trying to direct where those those individuals spend that debt.

    In the US, where the skills model is in place, social mobility is on the floor. Plus, there is a provable business link between labour market intelligence companies and Loans providers.

    Employer: Take on debt to come work for me, so you can pay off that debt.
    Me: Erm, no thanks.

  2. Agree 100% with the précis by Anon. Anon hit the bullseye with their astute observation of the proposed model representing the US model i.e onus of the cost of education and training falls on the trainee/learner.