Better jobs data key to levelling up

The labour market may be recovering but Neil Bachelor FRSA argues it still needs to be fixed and that one way of doing this is greater standardisation of how we communicate job vacancies and our work and learning history.

Although it is encouraging to see signs of a recovery within the labour market, its unfairness remains entrenched. A recent study by the Recruitment and Employment Confederation found that only 47% of Britons felt that the recruitment process was fair to jobseekers like them.

Labour market inefficiencies and inequalities are not new but sadly advances in technology and communication have failed to bring about much needed improvements. Progress has been made in as much as there is no longer a lack of work-related data (we are more measured, monitored and analysed than ever before). However, the data being shared is still too noisy and inconsistent to be a reliable, fair and efficient basis for decision-making.

The prime culprit remains the CV, which in some form is still required for most UK job applications. There are some conventions but their structure and vocabulary remain largely unpredictable and this makes evaluating them at any scale difficult and expensive for employers.

Technology has led to tools that automatically scan CVs and score candidates against a job specification but there are legitimate concerns about whether these opaque mechanisms are fair. Moreover, they have increased the burden on jobseekers to tweak their CVs to match advert keywords and second guess algorithms before they apply.

Inconsistencies and ambiguities in job adverts are also a significant problem. A study by Milkround in 2019 found that 68% of experienced workers wanted the wording of job adverts to be simplified. Three-quarters of graduates thought that job adverts could be deliberately ambiguous while 48% said they had even turned up to an interview unsure of what the job involved. The research also suggested a gender imbalance as women's confidence was particularly affected by technical jargon with 74% (compared to 61% of men) saying that unfamiliar acronyms made them feel under-qualified and less likely to apply.

Some employers argue that ambiguity is necessary to ‘sell’ roles and jargon useful in dissuading under-qualified candidates. If this is true though, it indicates an even broader problem with jobseekers struggling to find well-matched roles and both sides under-served by job boards who drive their revenue from premium listings; monetising jobseekers’ attention rather than pointing them directly to the most suitable vacancies.

What if instead there was a free and standardised way to communicate our work and learning history that was easy for both employers and computers to understand?

As well as helping individual employers and jobseekers, a standardised specification of jobs would hugely enhance the accuracy of labour market information. By being able to assess the skills of unemployed workers and labour market demand at a local, regional and national level, it would be much easier to identify and resource retraining pathways, ultimately leading to the more agile and resilient workforce we need.

There are already commercial providers analysing and aggregating job advert data at scale but access to this is expensive and is based on the same noisy adverts described above so there are questions about its accuracy. There are no alternatives though so these providers retain large public and social sector client bases; the data should probably be replicated and published openly on this basis alone.

Furthermore, the debate around levelling up has largely been about who should own and control reskilling budgets. Whoever this ends up being, the availability of free high-quality data allowing all sides to evaluate progress and outcomes should reduce at least some of the tension and hopefully allow greater co-operation between localised and centralised policymakers and providers.

It would also enable further innovation; a recent study into Nesta’s CareerTech programme concluding that: “The most common challenge encountered by innovators was the accessibility of labour market data which hindered the scale and pace of solution development”.  The ONS already publishes data from its Labour Force Survey and other sample-based studies but these are too expensive to scale and too slow to publish to provide the granularity and immediacy of data that we now need. Data derived from automated retrieval and analysis of standardised adverts would be far more cost effective.

So what should happen next? The large job boards and talent-tech providers do have the resources and expertise to implement the necessary changes but alas this would harm their business models. They are selling torches when we need more daylight. Instead, a public and social sector consortium should agree an open UK job specification framework based on existing taxonomies such as Schema.org, SOC-Extension, O*Net and ESCO. They should then pilot this, first within their own organisations and then throughout the public, social and private sectors. The labour market may be broken but with more co-operation and benevolent use of the technologies that we already have available, a fix isn’t far away.

Neil Bachelor is a Fellow and Director of Omnifolio, a Catalyst award-winning service that uses in-depth profiling and job-matching to help people change careers more easily.

Omnifolio is looking for data donors to complete a work questionnaire (~10-15 mins), which will help them complete validation of their profiling, matching and jobs library as part of a Catalyst funded project. In return, participants will receive personalised feedback about their strengths, motivations, behavioural styles and, for anyone considering their options, a list of job matches. If you’d like to take part, please register here.