This special issue of the Review has brought examples of the development of social protection policies and programmes that show whether, and if so how, reference to foreign experience has been made. It has sought to ask to what extent a learning process has occurred and to what extent policy-making in any one country can profit from an examination of proposals and solutions developed in another. Whilst doing this, it has paid attention to the role of international organizations, including the International Social Security Association (ISSA), in order to see both what their role has been and what it might be.

The case studies presented — namely, cost management in hospitals, the integration of disabled people into work, the financing of long-term care of the frail elderly and the provision of social pensions — provide support for the proposition that, when countries face common challenges or problems, not only do they tend to develop somewhat similar solutions, they also do look around to see what others facing such challenges or problems are doing. However, this is but a simplistic description of what really goes on. Pointing to what is being or has been done elsewhere might be one of the levers used by policy-makers to drive their own projects forward, but the details of what is being done elsewhere might be of much less concern to them. As the discussion of financing long-term care shows, it was useful for policy-makers in Japan to be able to refer to a successfully established, social insurance-based system in another country to support their own proposals, but they were not interested in the mechanics of the German system, merely its existence. Moreover, as the discussion of social pension systems shows, policy-makers in the various countries of southern Africa might have faced very similar challenges or problems, and they might have adopted rather similar solutions to these, but it is only recently that they, or their advisers, have started participating in any forum where ideas about social pensions are exchanged and discussed. Most policies are crafted for domestic political needs and for domestic consumption. The policies of other jurisdictions are rarely of great import. If policy-makers are conscious of them, they tend to refer to them sparingly and selectively, and they might even attribute to them characteristics, or give motives for their introduction, that they do not have. As the discussion of social pensions illustrates, policy analysts as well as policy-makers can err in the same way.

This is not to suggest that policy-making is seldom evidence-based and that learning rarely occurs. Policy-making is an iterative process that involves the need for improvement. Research can assist in this process and experience from abroad might form part of the evidence-base. The discussion of the development of policies to enhance the participation of people with disabilities in the labour market, and the discussion of systems to provide for greater control of inpatient treatment costs, show that approaches involving information gathering and analysis and an exchange of experience can both be recognized as beneficial and assist in taking programmes forward.

The iterative nature of policy-making reflects the fact that success is not always achieved and seldom at the first attempt. Mistakes can be and are made, and mistakes are to be learnt from. Learning from one's own mistakes is often painful; learning from the mistakes of others can be advantageous. Interestingly, the contribution that concentrates most on mistakes and learning from these — the discussion of policies to integrate disabled people into the labour market — uses examples of learning across time rather than learning across countries. Indeed, in none of the contributions is there the suggestion that policy-makers in any one country decided not to do something because they had seen that something failed elsewhere. This might be because, and as is well known with respect to pharmaceutical trials, “non-results” are seldom reported. It might also be because policy-makers engage in learning not so much as an aid in deciding between options but more as a means of confirming a particular approach or of seeing how to adapt and improve upon a measure that is already in operation.

Moreover, and as all four case studies make clear, even when elements of an approach that has been adopted elsewhere are of interest, that approach and its elements have to be tailored to national circumstances and to respect national contexts. The reasons for this are twofold. First, policies might well be more effective if they build on existing infrastructures. The way the Japanese long-term care insurance (LTCI) was built around the municipalities as deliverers, whilst in Germany it was built around the sickness funds, illustrates the virtue of this. Second, the process of adaptation enables domestic actors to be involved and to develop a sense of ownership that is vital for success. The way in which Diagnosis Related Groups (DRGs) were adapted to meet country-specific and system-specific needs in Germany, Switzerland and the Netherlands, rather than being “bought off-the-shelf”, was a consequence of a recognition that to adapt not only was more efficient but also more politically acceptable.

That policy-makers have to be cognizant of national circumstances could be taken to imply that the scope for reform based upon learning from abroad is limited. Indeed, two of the contributions suggest that all that can be learnt are technical elements of policy, and to expect more from learning across borders is over-ambitious. The examples of successful learning were those of DRGs and of special homes for people suffering from dementia.

However, if the term “learning” is given a wider meaning, a more positive conclusion can be drawn. The apparently successful introduction of an LTCI system in Japan provided a reform-inclined government in the Republic of Korea with an example to follow, even if the system the latter introduced was more modest. Thus, the suggestion that looking at other countries' experiences can raise the quality of policy-making is not to be dismissed. The formulation of country-specific DRGs profited from the ability of those concerned to look at how others had operationalized such a system. Equally, the information exchanges facilitated and promoted by the Organisation for Economic Co-operation and Development (OECD) had positive benefits for recipients of the relevant reports and for participants in the various meeting that surrounded their production. This, at least is the impression given by the discussion of policies to raise the labour force participation of people with disabilities. The latter contribution makes considerable play of the term “collaborative improvisation”, although its authors admit that it was not the first to describe policy learning in such a way. All the same, its analysis is plausible and illustrates what can and cannot be achieved.

Three of the contributions refer to the role international organizations can play or have played in showing how policies might be constructed, modified or implemented. The exception is the contribution discussing the introduction of LTCI, which could be because financing care of the frail elderly is a subject which the OECD — to which the governments of each of the three countries studied belong — has so far touched upon only tangentially. Perhaps, given the affiliation of the authors of this contribution, it is not surprising that the discussion of policies to improve the labour force participation of people with disabilities casts the OECD in a positive light. Other contributions, published elsewhere by other authors with different affiliations and mainly referring to other policy areas, have presented other international organizations in a similarly positive fashion. International organizations see themselves not only as clearing houses but also as policy entrepreneurs. They have “ideas” that they wish to disseminate. Ideas, of course, have to be argued for and defended. The contribution on social pensions makes clear that many argue that the introduction of such a benefit could satisfy both the immediate goal of alleviating poverty and the longer term goal of developing a wider system of social protection. Nonetheless, there has not yet been convincing evidence produced to show that social pensions are the most effective means of achieving the multiplicity of objectives claimed for them. Nor has any “entrepreneur”, institutional or otherwise, appeared who possesses full knowledge and understanding of all aspects of how such pensions may be financed, designed and delivered, or how they may be integrated coherently into a broader national social protection strategy.

This leads back to the point made in the introductory contribution. Learning has to be undertaken with a sense of adventure and a willingness to embrace new ways of thinking. There are elements of adventure embedded in each of the case studies. It depends upon both the teachers and the learners whether the opportunities offered by the learning process are taken up. International organizations, including the ISSA, certainly have a role to play here.