Skip to main content


Dipping into the Political Economy of Adult Learning Systems

I’m guessing that most people who read this blog will be familiar with Esping-Andersen’s typology of welfare regimes. The distinction between liberal, conservative and social democratic welfare regimes has provided more than one generation of researchers with a tool for investigating the ways in which structural differences in the relations between market and state affect various outcomes of interest – such as loneliness or social exclusion or active ageing. Although the typology has been widely used in multi-level analyses that try to tease apart the influence of important characteristics of individuals and features of the society in which they live, there is one major policy area that seems to have fallen by the wayside, and that is education.

Here the work of Hall and Soskice has been important in providing an alternative classification of modern (capitalist) economies that gives a central place to the institutional connections between education and training and the workings of a modern economy. One of the dimensions of distinction between liberal and coordinated market regimes is the approach taken to ‘human capital formation’, and in particular the specificity of the skills that are taught as a stepping stone to entering the workforce. This has led to a burgeoning literature – especially in continental Europe – on skills formation regimes, an idea which has provided researchers with a different lens through which to assess the importance of structural differences between societies. At one end of the spectrum there are so-called liberal skills regimes with little government or employer involvement beyond on-the-job training; and at the other end are collective skills regimes with high involvement of employers and unions in governance and financing of skill formation.

Recently my attention was caught by a number of reports and papers that looked for different ways of applying these (and similar) ideas to Adult Education or Lifelong Learning. What is the relative contribution of structural and individual characteristics in explaining participation and non-participation in adult learning? For the work that I’m doing at the moment, the importance of this question lies in the fact that age is a personal/individual characteristic that is strongly correlated with non-participation in job-related learning and training – and this correlation seems to be pretty much universal in high-income countries. One way of thinking about this is to ask how different structural features might mitigate or magnify the influence of individual characteristics. Dämmrich, for example, uses multi-level analyses to show that in Nordic countries male/female differences in participation are less than elsewhere (i.e. females have less ‘advantage’ in these countries than others, where their participation is higher). This particular paper unfortunately has nothing to say about age.

If the question of age is put to one side, what’s left are matters of more general or theoretical interest. One of the most sophisticated of recent attempts to deal with the question of structure comes from Richard Desjardins who has written several reports for the OECD on PIAAC data; and this was my introduction to ‘the emerging field of the political economy of adult learning systems’. His starting point is the existing typologies adapted from the work of Esping-Andersen and Hall and Soskice and their successors (they set the ball rolling as it were), and he ends up by developing a typology specifically for ‘adult learning systems’ (broadly conceived so as to include labour market policies, welfare structures etc., as well as the particular configurations of adult learning institutions).  

The main rationale for taking this route is that pre-existing typologies developed for different purposes do not line up very well with data on the overall level and distribution of participation in adult learning. How useful, for example, is the typology of skills formation regimes? One major feature of the data that seems to present a problem for this typology is the narrowing of inequalities in overall participation between various socially disadvantaged versus advantaged groups (i.e. young/old, low-SES/high-SES, low educated/high educated) in most countries including the market-led regimes such as US and the UK.  Another aspect of the data that he highlights is the poor correlation between total levels of welfare spending and the effectiveness of the ‘adult learning system’ in extending organized adult learning opportunities to the most disadvantaged. The correlation seems clear at the highest (the Scandinavian countries) and lowest levels of participation, but there is a bunch of countries with middling performance where it seems to be rather weak. “Austria, Belgium, France and Germany feature high overall levels of welfare state expenditures but lag behind Canada, Ireland, the UK and US in terms of extending organized adult learning opportunities to the most disadvantaged, even though the latter are below average spenders on all types of welfare programmes.”  

The solution to the theoretical problem (how to characterise the features of ‘adult learning systems’ that are effective in promoting overall participation and in extending to the most disadvantaged) is to include what Desjardin calls proximal features. These include educational structures that are ‘open’ to non-traditional learners and flexible in terms of qualifications suited to their needs, active labour market policies, and targeting of socially disadvantaged individuals. 

Take openness, for example. As a metric for success in this matter, Desjardins has data on the proportion of adults who attained their highest qualification as non-traditional learners, and it is striking how well some of the liberal economies perform (UK, US, Canada), not so well as the Scandinavian countries to be sure, but better than the rest of Europe. A metric is not the same as a mechanism of course, and for explanation Desjardins looks (inter alia) to the ways in which qualification systems work.  Another proximal feature that he highlights are active labour market policies.  The relationship with adult learning is not straightforward in this case. Not all spending on active labour policies translates into more effective adult learning systems, however. Without an open and flexible adult learning system, spending on active labour policies is unlikely to promote increased participation in adult learning. The point is perhaps obvious, but it serves to emphasise the importance of linkages and connections between different parts of the ‘system’. 

A final point for readers who might be sceptical about the UK’s showing in Desjardins’ data.  We are, after all, talking about an ‘adult learning system’ that was judged to be far from effective by a House of Commons Select Committee in 2020 (and a ‘skills formation regime’ that was found wanting in the 2006 Leitch review). Even though the Desjardins paper was published in 2020, the actual PIAAC data comes from 2012. It would be interesting to see how changes over the last 10 years would affect the analysis, not just the UK’s position, but the ‘model’ Desjardins is exploring. Is it stable over time?  As for the place of age in this analysis, another look at this particular form of inequality would seem to be in order.

About the Authors:

Kenneth Howse is a Senior Research Fellow at the Oxford Institute of Population Ageing. He is also a key member of The Oxford Programme on Fertility, Education and the Environment (OxFEE). 


Opinions of the blogger is their own and not endorsed by the Institute

Comments Welcome: We welcome your comments on this or any of the Institute's blog posts. Please feel free to email comments to be posted on your behalf to or use the Disqus facility linked below.