“One in eight Germans lives in poverty!” was the heading in Bild am Sonntag for a recent interview with Federal Minister of Labour Olaf Scholz. The Minister stated that according to the EU poverty definition, a single who earns less than sixty percent of the medium income is poor and that 13 percent fit this definition. The same line is taken by a study of the German Institute for Economic Research (DIW), according to which the share of Germans considered to be middle class declined by 3.8 percentage points between 2002 to 2005. And a study by the Institute for Labour and Qualification at the University of Duisburg-Essen asserts that the percentage of low wage earners in Germany (22 percent) now nearly matches that of the US (25 percent). Is there really a new poverty in Germany and are we losing our middle class?

Maybe. There are indicators that point to more inequality. The share of gross wages in the national income has decreased, and the spread of gross wages has increased in Germany, as in other industrial countries, although the spread is not as large as in Anglo-Saxon countries nor has it grown quite so fast as in most of them. The increasing inequality in market incomes is the result of the low wage competition from the former communist countries of the world. For many years now, I have been calling attention to the threat to social coherence in western Europe and to the necessity to reform the welfare state.

Nevertheless, I consider the cries of alarm that we are currently hearing in the media to be misguided and exaggerated. The public discussion suffers from a confusion of ideas, it ignores the existence of the German welfare state, makes use of outdated numbers and has allowed itself to be misled by a statistical artefact. Let’s look at each one in turn.

First of all, it is simply not true that one in eight Germans are poor. If this were the case, the German welfare state would be failing in its duties, defined in the Constitution and in the twelfth volume of the German Social Security Code, of safeguarding the dignity of its citizens through its social services. And indeed, in the government’s recent poverty report we find no such statement on the amount of poverty. The report refers to the “poverty risk” instead of poverty itself, and this is an important difference. In order not to stir up emotions, one should keep to the official concepts as defined by the Federal Statistical Office in its report on poverty and conditions of life (Armut und Lebensbedingungen) which is fully compatible with the OECD definitions. Here the “poverty risk” begins at 60 percent of the median, “relative income poverty” at 50 percent and “poverty” at 40 percent. The EU also does not say that one is poor at less than 60 percent but only “at risk of poverty”, as can be read in its reports on the social situation in the EU.

Hardly anyone with a legal residence in Germany is poor. Social welfare and ALG II (fixed welfare for the employable population) ensure, assuming normal housing costs, that income amounts to 54 percent of median earnings. Measured in terms of the relative reduction of the share of people at risk of poverty, Germany, after Scandinavia, is one of the most generous welfare states in Europe if not in the whole world. The percentage of those threatened by poverty – 26 percent – before social transfers (excluding pension payments) is reduced to 13 percent as a result of state social services in Germany, a decrease of exactly 50 percent. On average for the entire EU-25, state transfer payments have only lowered the portion of the population below the poverty risk line by 38.5 percent (from 26 percent to 16 percent). To be sure, about 4 percent of the German population is poor according to the official definition. But this has always been the hard core of those who for various reasons do not take advantage of the money offered by the welfare state and need to be helped by other measures.

In 2005, the year on which the government’s latest poverty report is based, the average monthly Hartz-IV benefit claim of a one person household stood at 700 euros, of which 360 euros went for housing and heating costs, according to OECD information. Free health insurance worth some 200 euros has not even been included. The poverty line however in that year was set at 520 euros. This clearly shows that ALG II recipients in Germany are not poor according to the standard international definition. They of course face the risk of poverty, since the corresponding limit was 781 euros. A two-person household and certainly large families are even relatively further above the poverty levels that apply to them because the additional grants that the German welfare state gives to additional persons in a household are larger in percentage terms than the value of their needs that are assumed when calculating the poverty line. Families are also not poor in Germany.

Because Germany has a welfare state, poverty is also not indicated when the portion of low-wage earners in Germany, with an hourly wage of less than two thirds of the median, approximates that in the US, because wages and income are not the same thing. The fact that the spread of gross wages is similarly high in both countries does not surprise economists: This is explained by the factor price equalisation theorem which can only be circumvented at the expense of mass unemployment. In the case of countries that trade with each other and between which capital can flow freely, wages must adjust for any given qualification level so that wage differences only result from wider gaps in education. But even these differences hardly exist, as the PISA tests confirm for Germany and the US. Both countries rank at the top in terms of the extent of the gaps in education among 15 year-old pupils. The tripartite educational system that has been proven to be a hindrance to social mobility is Germany’s real social problem and not poverty in itself. Poverty can be measured only in terms of income, which in Germany is much more evenly distributed, thanks to the welfare state, than in America. Germany’s low-wage earners receive ALG II that has been increased to such a great extent that virtually all recipients are well above the poverty line. They are even farther above the poverty line than unemployed ALG II recipients, who as already mentioned are at about 54 percent of median income, because the increases not only go to the ALG-II level but clearly above this. A low-wage earner who in 2005 worked full time for only four euros an hour had, assuming normal housing and heating costs, a net income of about 910 euros and was thus clearly above poverty-risk level of 781 euros. There is thus no cause for alarm on this point either!

However, how is it that according to the DIW analysis the share of the middle class in the German population (defined as between 150–70 percent of the median) decreased by almost four percentage points from 2002 to 2005, and that according to data from the Socioeconomic Panel the portion of those facing the risk of poverty increased by two percentage points? For a period of only three years, these are enormous changes; much more than could be explained by globalisation.

A portion of this increase is due to the special effect that appeared immediately after the Hartz reforms in the year of the report, 2005. The Hartz IV reform meant that about two million Germans were reclassified from unemployment assistance recipients to social welfare recipients and that wage subsidies were paid to low-wage earners in the context of ALG II. Immediately after the introduction of the new welfare system, this led in fact to an increase in the share of those facing a poverty risk because of the high unemployment, which was the highest in the history of the Federal Republic. In the meantime, however, the Hartz IV reforms have taken effect and have brought about a miracle on the German labour market. In western Germany alone, the recent economic upswing has created at least 1.1 million more jobs than could have been expected from an extrapolation based on previous economic upturn patterns. The gain in jobs together with wage subsidies have made considerable contributions to the decline in the risk of poverty and to a stabilising of the middle class – but this is not yet evident in the statistics of 2005. The government’s poverty report was already dated before it was written. Because it takes no account of the labour market successes, it is impossible to use this report as proof of increasing poverty in Germany, deficits in the Hartz IV legislation or even the need for minimum wages. The latter would be especially absurd, since minimum wages force an even larger portion of the lesser qualified into unemployment.

Doubts about the interpretation of the numbers are also justified because the development presented is not based on a factor such as per-capita income but on what statisticians call income based on needs-adjusted equivalence scales. Here it is assumed that a couple that separates is considered to become poorer unless each partner’s income increases by a third. This is not implausible but it implies that the redistribution measurement is more an extension of social preferences than economically induced support-shortfalls. The principle that usually applies in statistics of separating facts and value judgements is being violated by the poverty statistics.

Because of the weighting of the equivalence scale, a reduction in household size but with the same per-capita income leads in itself to a reduction in equivalence-scale income – and thus to a larger share of those facing a poverty risk and falling out of the middle class. A decrease in household sizes has indeed taken place to a large extent in Germany: According to the OECD, the share of single parent households grew four times faster in Germany than the OECD average between 1995 and 2005. 80 percent of the growth of the calculated inequality in this time period can be explained by the changed age and household structures. From 2002 to 2005, the period on which the poverty report is based, the number of single-person households grew by a considerable 3.3 percent.

Here, the welfare state provides for the additional needs of the eroded families by granting additional benefits. But it does not do this to the extent that statisticians who devised the equivalent scales would prefer. The German welfare state gives a couple living on ALG-II that decides to separate only an increase in the standard benefit of eleven percent instead of a third more as the equivalence scale would require.

The increase in the (computational) poverty risk thus has little to do with greater inequality in the wage structure or a deficit in the structure of the welfare state. Public opinion gets worked up over the inequities of the market economy – and has been duped by a statistical artefact.

The whole thing becomes a farce if we consider that unemployment benefits and to a much greater extent the current ALG II offer a strong stimulus to couples to live separately. If one partner earns well and the other is long-term unemployed, the unemployed partner can only claim benefits if he/she does not formally cohabit with the well-earning partner in a “household unit”. If they both form a joint household, the benefits are cancelled. State aid thus becomes a separation bonus, the state only paying on the condition that the couple does not marry or formally cohabitate. No wonder that many young people chose not to declare cohabitation or separate if they once did. The result: In the first year of ALG II, the number of single person households increased in Germany by one percent and in the second year by five percent.

Although the additional welfare benefits make it economically possible for people to live alone, the statistics then indicate that there is a higher risk of poverty. If the partners live together, they belong to the middle class and are not regarded as in risk of poverty. If they declare that they have separate households in order to get more money from the state, the unemployed partner is now suddenly considered to be in risk of poverty. (According to a court decision, a separation of table and bed within a home is sufficient for declaring separately.)

It is obvious that when the needs-adjusted equivalence scales are used to calculate redistribution measures, they do not lead to sensible results if this causes families to erode and the state supports this erosion financially, as is the case in Germany. Behavioural changes that people adopt to improve their economic situation are registered as a worsening of their living situation in the statistics and lead politicians to sound the poverty alarm. This is needs-adjusted nonsense, nothing more.

Hans-Werner Sinn
Professor of Economics and Finance, University of Munich President of the Ifo Institute

Slightly shortened version of „Der bedarfsgewichtete Käse und die neue Armut“, Ifo Schnelldienst 61, 2008, p. 14-16; additionally published in WirtschaftsWoche, no. 22, May 26, 2008, p. 48f.