Farewell to the Swedish Model? Reflections on the Beveridge Social Security Conference

David A Preston
General Manager, Social Policy Agency

In September 1992 the University of York hosted a Social Security Conference to mark the 50th Anniversary of the Beveridge Report and its proposals for a National Insurance System for the United Kingdom. The large numbers of papers presented meant that the discussion was multi-faceted. However, certain key themes emerged. For British participants the central issue in the deliberation was the extent to which the Beveridge "National Insurance" model was still relevant to Britain in the closing stages of the twentieth century. For the rest of us the issues were much wider. For those from developed countries they focused on where any of our social security and social insurance systems were going in the charged conditions developed countries were not facing. For those coming from developing or former communist countries the issues centred on what types of social security system were appropriate to their circumstances. This paper focuses mainly on the issues of social security systems in the existing group of developed countries.

What has changed?

It was clear to most conference participants that conditions in most developed countries were now very different from those when the existing social security and social insurance schemes were set up. Indeed the economic and social situation of most countries is now very different from that of even twenty years ago. It also seemed that most countries were facing significant problems in budgeting for social security or dealing with potential shortfalls in social insurance funds. These problems tended to reflect social and demographic factors which were common to most developed countries, as well as the impact of a slower trend in economic growth. The significant background factors are outlined below.

• Developed country populations were now ageing, and future projections were for a further rise in the proportion of elderly people. This demographic ageing was expected to place an increasing strain on pension and health systems.

• Unemployment was running at high levels in most countries. Apart from the impact on affected people, this was also placing a strain on social security resources.

• Disparities in earned incomes seemed to be increasing in many industrial countries.

• Social changes affecting family and employment structures were also widespread. The model of "male breadwinner/female career" family with children was now less prevalent. Family structure and roles had been significantly affected by the following factors:

(a) A rise in the proportion of married women working, leading to more two income households.

(b) A rise in the proportion of sole parent households, many of whom had become a new dependency group for the welfare state.

• A "new poverty" seemed to have emerged in many countries – or at least become more socially visible. The elderly were now less likely to be poor – a success story of the welfare state – but comparative poverty was now found more frequently amongst low income couples with children, sole parents, and in some cases unemployed young adults. The new poverty also seemed to disproportionately impact on ethnic minorities and recent immigrant groups.

• Government in a number of countries had become concerned about work incentives and poverty traps in benefit systems.

• Social and political philosophies had been shifting in a number of countries and calling into question established views on the appropriate role of the state in providing income security.

All these factors appeared to add up to a situation where the appropriateness and cost effectiveness of existing social security systems was under question. Hence the stability or direction of development of social security or social insurance structure was less certain than seemed to be the case a generation ago. Across countries, a pattern of grappling with common problems was more evident than any uniform direction of system development.

The convergence thesis?

One way of looking at the issue is to contrast recent developments with what might be called the "convergence thesis" on social security. In its most simplified form the convergence thesis postulated that, as economies develop along modern lines, certain common characteristics and patterns emerge. Increasingly, developed countries would become more and more alike in their economic and social structures. While there might be leaders and laggards in the process, the convergence theory postulated that social security systems in developed countries would also become increasingly similar.

The most eloquent exposition of an earlier version of the convergence thesis was put forward by Marshall, an early academic populariser of the Beveridge report¹. His historically based analysis suggested that western countries had gone through three stages of definition of human rights. In the first stage the focus was on the civil rights of individuals, the core freedoms. A second stage saw the focus on political rights, such as the right to vote. The third and final stage saw a focus on economic rights, including social security and healthcare, as the enfranchised population used their voting power to reshape society in a preferred mould.

Underlying the social security dimension of the convergence thesis as it was subsequently popularised were some analytical views on social, economic, and political changes which had been occurring in western societies.

Socially, the process of economic development and accompanying urbanisation was seen as eroding the extended family and clan structure of more traditional societies. Until comparatively modern times these structures had provided the main source of economic security for individuals as they passed through their life cycle, or faced unexpected setbacks. Modernisation and the accompanying erosion of these mutual support networks created a vacuum which private insurance could only partly fill.

Secondly, economic development itself produced both the resources to finance social security systems and demonstrated the limits that could be achieved by solely private provision. Greater wealth created greater possibility for the public authorities to acquire and spend revenues. At the same time, the private commercial sector could solve only some of the needs of individuals and nuclear families for an income support system which protected against the contingencies of life or expected life cycle changes such as ageing. The key problem for a voluntary private insurance model was that of adverse selection. Those most likely to need assistance – the poor, the disadvantaged, and the feckless – were least likely to seek, and least able to obtain, adequate cover from the commercial sector. Hence there was a need for some other source of assistance for significant groups in the population, with the public sector being the most obvious source.

Finally, there was the political point made by Marshall, that newly enfranchised majorities in western democracies would ultimately use their voting power to gain economic rights which they saw as corresponding to their needs. Social security was just such a case.

Twenty years ago the most popular version of the convergence model seemed to be the idea that we were all becoming like Sweden in the 1960s. For social security this meant a system of high taxes or social insurance contribution which funded high universal benefits and social services. Overall, the social security system would be characterised by egalitarian and fully employed populations, with income security at high levels underwritten by the social security system.

Strains on the convergence thesis

Whether the convergence thesis or some variant of it still had any predictive validity for social security systems was one of the issues at the conference. Again, there were at least two separate debates at the conferences which only partly overlapped.

The first debate was an intra-European one over whether the separate social security systems of European Community members would converge. The leading convergence proponent was Julian le Grand. He argued that, of necessity, the Community systems would become similar as without this, a common labour market could not work effectively. The outcome of non convergence would be the hindering of economically efficient migration (to where jobs were available), and the encouragement of economically inefficient migration (to take advantage of the most generous welfare systems). However, a newspaper headline which accompanied the Conference and the ongoing debate on the Maastricht Treaty indicated that the UK Minister of Social Security seemed to be less than enthusiastic about this particular convergence thesis, and indicated a view that separate systems would be likely to continue for the foreseeable future.

The wider debate, which was of more interest to participants outside the European Community, was whether convergence had actually happened, or was still in evidence, given the diverse origins of different systems and the increasing problems of financing them. The starting point for this debate were the proposals of Esping Andersen advanced in the "Three Worlds of Welfare Capitalism". Esping Andersen had argued that developed economies had evolved three alternative models of social security rather than simply one. His framework classified welfare states into the following three groups.

1. Social Democratic (the Swedish or citizens' rights model with high universal payments).

2. Conservative (based on the Bismark derived contributory social insurance principle with earnings related benefits and pensions).

3. Reluctant (the American or public charity model with limited, conditional, and highly targeted assistance).

The third group, incidentally, was held by Esping Andersen to include Australia, New Zealand and the United Kingdom.

In the discussion which followed it was clear that the Esping Andersen thesis was regarded as over simplified. Many disputed that the US system was the same as the British, Australian and NZ cases. In addition, new systems such as those of Korea, Taiwan, Singapore or Hong Kong provided a fresh range of alternative models. This view led to the conclusion that there were perhaps five or six alternative social security models now operating. At the extreme it could even be argued that every country was a separate case.

Underlying this debate was the recognition that the separate dimensions of social security systems could be combined in many ways to give a large number of varied systems.

• Firstly, sources of funding can be general taxes, earnings related contribution from employees and employers, flat rate national insurance premiums, or dedicated taxes, or a combination of these.

• Secondly, the benefits or pensions payable from the scheme may be earnings or contribution related, flat rate payments, or income or asset tested payments.

• Thirdly, levels of payment may be high, medium or low in relation to characteristic earnings.

• Fourthly, the nature of the conditionality attached to payment could vary enormously. For example, public pensions may or may not require actual retirement, while definitions of "invalidity" were extremely diverse.

• And finally there was a range of units of assessment on entitlement issues – e.g. whether payments related to individuals, couples, nuclear families, or some other aggregation.

The possible combinations are quite large. However, it also tends to be the case that some groupings fit together more easily than others – for example earnings related contribution schemes are normally associated with some form of earnings related benefits or pensions, while highly targeted payment systems are more commonly funded from general taxation.

Summary

The patterns evident at the Beveridge conference were more a stocktaking of systems in flux than an identification of a common social security destiny.

It is no longer clear (if it ever was) that modern economies are experiencing a convergence in their social security systems. The most consistent theme across countries was a pattern of rationalisation and cost containment strategies within existing systems rather than a conscious move to a common model. Certainly, if any convergence re-emerges in the short to medium term it seems unlikely to be towards the "Swedish model". Even the Swedes appear to be trimming back on their own system.

For the longer term, the convergence issues are still open. Certain factors may lead to a revived convergence pattern amongst some states. These include the expanded European community, migration linked bilateral and multilateral social security agreements, and the politics of emulation in affluent societies. At the same time economic forces may impose a "convergence of limitation". International mobility of production, enterprises and highly skilled labour could limit the tax levels which governments can set within their own national economic and social boundaries, and thus impose limits on sustainable social security systems. Also, social security is only one of a variety of competing public expenditures within the public choice matrix.

¹ Ref: Peter Baldwin "Beveridge in the Longue Duree" Conference Paper. Volume A of Conference Proceedings.