June 26, 2005

How high is total Swedish unemployment?

By Tino

This might not seem like a difficult question. After all, the unemployment number is readily available. It takes a few seconds to check up the American unemployment number, 5.1% in May. Sure there are speculations about discouraged workers leaving the workforce, but overall this is a reliable figure, useful for international and cross time comparison.

Not so with Sweden. Yes, there is an official number, 5.2% in May 2005. But this figure is almost compleatly worthless. The true unemployment number is in fact closer to 20%, as shown below.

Sweden has several massive government programs that contain must of the unemployed, chiefly Early Retirement, Sick Leave, Labor Market Programs and Welfare. In addition, since students get a monthly stipend and loan, many unemployed continue to study when they can’t find work.


The right in Sweden has for some time complained about the government hiding the unemployment in ‘programs’. But the question only came to the public knowledge when LO, the immensely powerful blue-collar union had one of it’s own turn against them. Hans Karlsson, a leftwing heavyweight, concluded that true unemployment was more in the ballpark of 20-25%, not 5% as the government was claiming. To make things even worse this guy happened to have the name as the Minister for Labor!

A few weeks later another labor movement veteran actually become a whistle blower, resigning and exposing that LOs left wing research institute had pressured him not to include the estimated true unemployment figure in a study. Embarrassing stuff indeed.

For a century the Social Democratic Labor Party parties and their allies in the Unions had Full Employment as one of their two ideological pillars (the other being an equal distribution of income). To now admit they have failed in this would be a huge defeat.

Furthermore, the party won the last two elections on the platform that cutting taxes would jeopardize underfunded Social Services. If Swedes were to conclude much of their taxes doesn’t go to welfare, but to support working age adults, their willingness to continue paying them might be reduced.

One key to understand this figure is that we have seen in the past few years is that the already high rate of sick leave has exploded. Yes, you wouldn’t guess it if you saw us, but Sweden has the sickest population in the world. Some 14% of the adult Swedish population does not work due to reported sickens. 540 000 were early retired (officially for health reasons) December 2004, fully 10% of the adult population. In addition more than 5% of the working population is on sick leave any given day.

Estimating the true unemployment rate

I will try to calculate the figure using data from the Swedish Statistical Agency (SCB) Labor survey, first quarter 2005. The calculations and adjustments are too messy to show here, but will be given by request.

Two notes. I have adjusted their Sick Leave number. Also all figures are from SCB, except the number only on welfare which I got from Tax researcher Dane Nordlings homepage. http://danne-nordling.blogspot.com/2005/05/hur-mnga-r-egentligen-arbetslsa.html

Table 1. First quarter of 2005, adjusted.

Population 16-64 - -- - - 5755.000

Not in laborforce - - - - - - 1.365.000
Early retired - - - - - - - - - (494.000)
Seek work - - - - - - - - - - - (150.000)
Get Welfare - - - - - - - - - - (84.000)

Labor Force - - - - - - - - - 4.391.000
Unemployed(5.6%) - - - - - (245.000)
unemply programs - - - - - (130.000)

Employed in Real job - - - - 4.016.000
Absent from work - - - - - - (554.000)
Of which on Sick leave- - - (216.000)

Actually work - - - - - - - 3.469.000

4.0 million (70% of adult population) in productive activity rate, 1.2 million (20%) living of welfare and Health or unemployment insurance alone.

Range of estimate: 8.5%-26%

So what is Swedish unemployment? The question hinges on who we include, and especially Sweden’s case especially on how much of early retirement and sick leave is hidden unemployment.

Even the Swedish governments acknowledges that the 130.000 in “Labor market political programs” simply unemployed, so we start from 8.5%.

For international comparisons we probably should not include the students and other’s who want work but can’t get it, since all countries have this category. But for policy matter in Sweden we should. Also including welfare recipients we get 13.2%.

The hard question is what to do with all the absentees. If we include all of them we end up with 26%. Note that we are not including absentees for any other reasons than Sick Leave and also excluding the underemployed.

Now this last figure is clearly too high. Many people who report sick are of course really too sick to work. It does show us the range, and is probably closer to the actual number than 8.5%.

It is anybodies guess how many who could go back to work if it paid to work. There is plenty of evidence that strong economic incentives to report sick matters. (I full account on this research another day).

Sweden had some 26 days lost per worker in 2002. In Canada, hardly a much sicker or slave driving nation, the corresponding figure (2004) were 7.5 days per full time worker lost every year due to sick leave (Statcan.ca). Britain had a similar figure, 7.2 days lost per worker and year in 2003 (CBI annual absence survey)

My Best Guess

As plausible comparisons I use the rates of early retirement in 1960 and the sick leave of slightly above Canada’s. Both would seem to exaggerate the number, we are for example much healthier now than in 1960.

Not counting the forced student, the underemployed or all other absentees the Swedish unemployment rate would be 19.8% fully 962.000 out of work of a workforce of 4.848.000 (add the latent jobless to get 22.5%).

Unemployed (245.) + Programs (130.) + Welfare (84.) = 459.000

Unemployment hidden in Early retirement (373.000) and Sick Leave (131.000) = 504.000

Total 962.000 out of work, Labor force 4.848.000.

In practice I think this 19.8% figure is somwehat comparable to rates for the US and other market economies. I am not sure however we should compare it to the “raw” figures of Germany or France, since they too hide unemployment in other welfare systems (though probably no country does as much as Sweden).

Next time I hope to write more detailed on how this high unemployment and welfare dependency figure has lead to chronically underfunded Social services.

By tino.sanadaji@gmail.com

Posted at June 26, 2005 01:03 AM



Good posts. One thing that is not often noted, I believe, is that despite the muddled Swedish economy, Sweden has a very intelligent population boosting one of the highest avg. IQs and within the population there are a lot of inventive types who seem to do quality work despite the incentives to work otherwise.

Not all countries have these advantages within their populations, and I therefore think that the Swedish economic model will produce different and poorer results if adopted by other countries.

Comment by larry at June 26, 2005 09:12 PM | Permalink

A good post that quite strikingly puts the finger on perhaps the biggest problem for swedes that do not want to see their country go down the drain.

However, as long as so few dare admit the real numbers concerning unemployment - little will happen.

Keep up the good work.

Comment by Sebastian Weil at June 28, 2005 02:26 PM | Permalink

Post a Comment

Remember Me?

(you may use HTML tags for style):

Note: You may have to reload to see your comment.

Trackback Pings

TrackBack URL for this entry: