The Swedish private unemployment insurance model

In recent years, new private unemployment insurance schemes administered by unions have developed rapidly in Sweden. In these schemes, union membership has for the first time been made obligatory. This development has at least three consequences. First, making membership mandatory can help unions combat the decline in membership rates.¹⁰⁴ Second, this shift could change unions’ incentives in relation to the public insurance system moving from support of high quality insurances towards unions being indifferent or supporting retrenchment. Third, this development is likely to exacerbate dualization trends in so much as these union schemes are likely to only cover labour market insiders due to smaller, sector-based, insurance pools.¹⁰⁵

For a long time, unemployment insurance has been synonymous with publicly provided unemployment insurance since it was first incorporated into the public social-insurance system during the early-to-mid nineteenth century. In fact, private provision has more or less been ruled out as an alternative as the risk of unemployment has been regarded as

103 Tomi Kyyrä, Pierpaolo Parrotta and Michael Rosholm: The Effect of Receiving Supplementary UI Benefits on Unemployment Duration, Department of Economics, Working Paper 09-1, pages 19-22; ISBN 9788778824042 (online) http://pure.au.dk/portal/files/4914/WP_09-1 (17.01.2013)

104 Scruggs, L. (2002). The Ghent System and Union Membership in Europe, 1970-1996. In Political Research Quarterly, 55(2): 275-297.

105 Clasen, Jochen, and Elke Viebrock, (2008), ―Voluntary Unemployment Insurance and Trade Union Membership: Investigating Connections in Denmark and Sweden‖ Journal of Social Policy, Vol. 37, No. 3, pages 433-452.

uninsurable by market providers. The existence of highly developed public insurance schemes has also more or less crowded out private alternatives.

It is surprising then to see a sharp rise in privately provided unemployment insurance in Sweden, a country with arguably one of the most generous public systems. One of the explanations for this development can be found in the continuous decline in the quality of the public unemployment insurance system.

The spread of private unemployment insurance is the result of unions deciding to supply their own supplementary insurance in response to the decline in the quality of the public system – a majority of workers are now covered by private insurance.

Unions in Sweden have argued for increasing the level of the unemployment benefit ceiling during the 1990s and 2000s.¹⁰⁶ However, when it reached a critical level and when the government introduced absolute cuts in the ceiling, unions shifted their attention to potential alternatives and began introducing a new and supplementary form of insurance. This insurance is tied to union membership. The effective tying of unemployment insurance to membership is an important policy innovation.

A lot has been written on the relationship between Ghent systems of unemployment insurance and union membership recruitment.¹⁰⁷ In that literature, it is argued that the unions’ position in the administration of the unemployment insurance functions as a recruitment device. Since the unemployment insurance is voluntary and administered by unions, workers are argued to have incentives to join unions as members. However, such ties are weak since there most often has been no requirement of insurance takers to join the unions as members.

The Ghent system and union membership. In most countries, the unemployment insurance is mandatory for all workers and administered by the state. However, some countries have retained the so-called Ghent system, which was more widespread in the early 20th century. The Ghent system can be characterized as a state-subsidized insurance with voluntary membership and union administration. Today, the Ghent system can be found in Denmark, Finland and Sweden. In Belgium, the unemployment insurance was made compulsory in 1944, but the unions still have a role in its administration through the payment of benefits.¹⁰⁸

106 Davidsson, J. B., (2011). Unions in Hard Times. Labour Market Politics in Western Europe: Two Patterns of Reform. PhD Thesis, European University Institute, Florence, Italy.

107 Kjellberg, Anders, (2006). ―The Swedish unemployment insurance – will the Ghent system survive?‖

Transfer: European Review of Labour and Research, Vol. 1, No. 6.

108 Vandaele, Kurt, (2006), ―A report from the homeland of the Ghent system: the relationship between unemployment and trade union membership in Belgium.‖ Transfer: European Review of Labour and Research, Vol. 4, No. 6.

The institutional variation that exists between countries in terms of unemployment insurance systems has been argued to account for differences in the rates of union membership in the workforce (union density), where Ghent systems produce higher union membership rates. Scruggs (2002) has shown a correlation between unemployment insurance institutions and union density.¹⁰⁹ The Ghent countries are not immune to the overarching trend of decline in union density; in fact they have seen quite important losses in recent years.¹¹⁰

However, nowadays unemployment insurance funds are no longer run by unions, but by agencies that are financially separate from but linked to unions¹¹¹. In addition, it is job agencies, run by the state, local authorities, or by private companies, that have the responsibility to ensure that benefit receivers live up to the demands for job search activity and acceptance of suitable jobs.

The rise of private union-run supplementary insurance provides for the first time a clear link between union membership and membership in the unemployment insurance. In a way, the rapid rise of such insurance in Sweden can be referred to as the creation of the first real Ghent system.

The diminishing quality of the public unemployment system is a general phenomenon in many countries, including Sweden as well. Therefore, all countries, with the exception of Belgium, have retained the Ghent system of unemployment insurance with a voluntary and union administered insurance. The other dimensions of the insurance have gone through considerable changes. Most importantly the quality of the insurance, in terms of its capacity to replace previous earnings, has been undermined.

For example, in Sweden, the change of the replacement rate has been minimal, from 90 per cent of previous earnings in the 1980s to 80 per cent in the period after the economic crisis of the 1990s. However, in the early 1990s, during the crisis years, the government decided to de-couple the ceiling in the insurance from wage increases. The sharp rise in unemployment made the government weary of not being able to control prospective increases in costs. The effect was a hollowing out of the earnings-related character of the insurance – today a worker with an average wage will only have a

109 There is quite an extensive literature that supports the correlation between Ghent systems and high union density (e.g. Western 1993; Blaschke 2000; Scruggs 2002). However, the causal link has been quite weak since trade union membership is never really tied to membership in the unemployment insurance. The causal argument is based on Olson’s (1965) theory of ―selective incentives‖, which holds that participation in collective action presupposes that some benefits are available only to active participants. If not, and if the outcome of collective action is a public good, there are too strong incentives to free ride.

110 Lind, Jens, (2009), ―The End of the Ghent System as Trade Union Recruitment Machinery?‖ Industrial Relations Journal, Vol. 40, No. 6, pages 510-523.

111 Davidsson, J. B., (2011). Unions in Hard Times. Labour Market Politics in Western Europe: Two Patterns of Reform. PhD Thesis, European University Institute, Florence, Italy.

replacement rate of about 50 per cent. This development represents a clear break with the Swedish universal model of welfare.¹¹²

The qualification requirements are also stricter in the private insurances than they are in the public system. The development of private insurance might also change the political dynamic in such a way as to decrease the support for universal public insurance. The universal system was built on the premise that it is essential to include the middle-class, or labour market insiders, in the system in order to safeguard political support¹¹³. If insiders become reliant on private insurance, and no longer need the public system, it is likely that their support for the public system would diminish, leading to lower quality public insurances. This in turn would produce negative effects for labour market outsiders as they are likely to be restricted to the public system for social protection. If the public system lowers quality by reducing the ceiling in the unemployment insurance, there will also be a larger gap to cover for the private insurance funds, which in turn could reinforce their need to offset risks, leading to less coverage and stricter qualification requirements.¹¹⁴

This political dynamic has already begun to emerge in Sweden. As mentioned above, the conservative coalition has both decreased the quality in the public insurance and publicly endorsed the development of private insurance. It is quite clear that they envision the public unemployment insurance to gradually be transformed into a sort of low-level flat-rate insurance making it necessary for those with average to high wages to insure themselves on the private market. This dynamic can be reinforced if the unions withdraw their support for a high-quality public insurance in favour of their own private insurances.¹¹⁵