Occupational data quality, coverage and availability has come on leaps and bounds over the last 20 to 30 years, and since the development of O*NET. O*NET marked a major in advance in several ways:

  • It combines a very comprehensive set of occupational descriptors with direct link to occupational classification
  • It is national and economy-wide in its coverage (for the USA)
  • Its core content model is both theoretically informed and validated
  • Access is through multiple routes ranging from job titles to occupational classifications (and groupings) plus key work level descriptors
  • It supports relatively easy cross-occupational analyses and comparisons
  • Cross-walking to other occupational classification systems and supports bringing together multiple occupational level datasets1

The O*NET dataset provides the core on which to build but it is important to recognise the limitations and the challenges they present to users and raises pointers as to which datasets could be added to enhance its value. The points which follow offers some insights into both areas.

  • Approximation: the task data and related data collected from job holders and through analysts are a summary of what is done and is not exhaustive. What is collected is the core and most important activities which is a great start but not the whole picture.
  • Individual focus: quite rightly the data are collected from individuals about the jobs they hold and so does not overtly reflect team activities and/or work processes which call for combined efforts from several individuals working in jobs across several occupations.
  • Proficiency and effectiveness: none of the descriptors help with viewing the minimum or optimum levels of proficiency and how effectiveness varies across job holders.
  • Historic: as with all surveys, the data are accurate for the time they are collected and for many occupations this is not an issue as they are not changing rapidly. O*NET does update occupations on a rolling basis covering around 10% of occupations on an annual basis. In addition, O*NET also seeks to cover emerging and new occupations and tasks, and those for specific sectors (e.g. the ‘green sector’). Combining O*NET with online job scrapings might be one way to enhance the data recognising the limitations of such data.
  • Individual potential: O*NET describes the work as done each day by job holders and so represents the demand for skills, knowledge and abilities required by specific occupations. These data do indicate the capacity and capability of job holders as can be expressed within the boundaries if the jobs they do. However, this will leave many areas unstated as regards what the individual can and might be able to do. This is of particular interest where so many economies lag in the development of sustainable ‘graduate level’ jobs, meaning many university graduates hold jobs which are well within their capability.
  • Professional standards: as labour markets develop, increasing numbers of occupations are covered by a professional body and carry with them the requirement for the licence to operate (practice). In the UK about one third of the labour market is either professional or associate professional level which means specific bodies control entry into an occupation and shape the content of the initial and subsequent professional training. And, by default, the professional bodies shape the content of the occupation. O*NET recognises this as regards entry but there is some work to be done to marry content with national and international professional bodies. This issue is also true for other externally, employed-based standards as you find in apprenticeship standards and more broadly in national occupational standards.
  • Non-task activity: the work content data (task through to detailed and intermediate work activities) all focus on specific work tasks and do not cover self-development or process improvement tasks across occupations. This capacity to improve is very important when seeking to map the resources and occupations deployed to make them happen.
  • Location: occupation content data can be tied to a geography through the location of an employer or business activity. Increasingly the location of a task can be very flexible and allows for work to be sub-contracted locally at a place of work or at many miles distant (e.g. software development can take place almost anywhere). As increasing interest is being shown on the impact of AI on occupations and how this does, and might vary across time and space, location data become more important.
  • Work quality: Using the content model and descriptors used by O*NET it is possible to piece together those jobs which will provide potentially a high level of satisfaction and high quality. Such analyses are indicative, and it would be very useful to combine work quality data with O*NET perhaps using the largescale European surveys on the topic.
  • Career progression: O*NET does develop a series of career clusters and career pathways, but these could be made more explicit and the careers function does allow for crosswalks to be viewed at the individual level. More work is required here to allow education and training providers to better understand and resource for the bridging and transition programmes required to update and shift potentially low demand skills sets to higher demand ones.
  • International occupations: Both O*NET and ESCO are used internationally with O*NET being used in over 140 countries with local adaptation. It would be useful to be able to identify those occupations descriptions which are seen as being largely common almost irrespective of geography and also highlight the regulatory differences which preclude which job holders can do what e.g. as in the medical and health professions around diagnosis and treatment (administering medicines and changing prescriptions).
  • Hybrid occupations: Being able to understand and map how skills and whole occupations emerge, change and develop and then finally being defined and captured as being “stable” and “owned” by a specific occupation. Knowing which occupations are those which are first to absorb and undertake “new work” would be very helpful with informing the impact of future technologies have upon occupations and their content.
  • Workplace experienced-based, vocational standards: While O*NET and related occupational datasets provide a profile of work requirements demand, they also describe sub-occupation level standards and could be used as a way of recording and potentially recognising work experience. Ideally this could be accredited in some form of prior learning developed whilst at work. This would be particularly helpful for those job holders in relatively low skilled and potential at risk occupations and who will be forced to make job and occupation changes.

Notes:

  1. National Research Council (1999) The Changing Nature of Work: Implications for Occupational Analysis. Washington DC, The National Academies Press. See pages 195-196