Subsidised jobs and more increased assistance for job seekers, as well as better employment training also seem to produce positive effects, the study found.

“As opposed to the latest review, our overall opinion is that the labour market policy is working fairly well. One possible explanation for this is that the interaction with the unemployment insurance has changed,” the government said in a statement released along with the initial findings of its Long-Term Survey (Långtidsutredningen) of the Swedish labour market.

Anders Forslund of the Institute for Labour Market Policy Evaluation (Institutet för arbetsmarknadspolitisk utvärdering – IFAU), one of the researchers participating in the Long-Term Survey, thinks that it will be difficult for Sweden to return to the low equilibrium unemployment rates similar to those recorded prior to 1990.

“It’s hard to find any large flaws in today’s labour market policy, so reaching a rate of 2 to 3 percent seems unlikely,” he told the TT news agency.

He attributed Sweden’s earlier low unemployment rates to the fact that unemployment was never allowed to grow during difficult times, for example through currency devaluations.

“We avoided large increases in unemployment until we exhausted our possibilities to live beyond our means, in the early nineties,” said Forslund, who pointed out that very few people in Sweden were classified as long-term unemployed until the financial crisis of the early 1990s.

According to Forslund, during the nineties it also became easier to be unemployed for a long time, whilst receiving unemployment benefits.

Demands for an improved and extended employment training have been put forth in current political debates, but Forslund warned that extra training is no miracle cure.

He acknowledges, however, that employment training could perhaps comprise more than it does today.

He points out that results from employment training in the 1980s and 2000s were good, but results in the 1990s were less stellar. This was partly because the training entitled participants to new periods of unemployment benefits.

”Many people participated mainly to get renewed benefits, not to acquire useful qualifications for a new job,” he said.

The large number of jobless Swedes in the 1990s also worsened the quality of the training.

When it comes to subsidized employment, such as the government’s “new-start jobs” (nystartsjobb), the researchers conclude that the measures generally work well, but also emphasize the importance of making sure the subsidies reach those who really need them, so that positions that the employer would have filled anyway aren’t crowded out.

As for employment programmes, such as the job and development guarantee for young people and long-term unemployed, or efforts with job coaches, Forslund concluded that these measures are difficult to evaluate.

The Long-Term Survey is a study of the Swedish labour market, as well as an evaluation of the effects of Swedish labour market policies. The results of the study are published yearly by the Institute for Labour Market Policy Evaluation (Institutet för arbetsmarknadspolitisk utvärdering – IFAU).

The main report is due to be published in the spring of 2011.