Official poverty rates in the European Union are usually based on disposable household income after taxes and public cash transfers (e.g. child benefits) collected from nationally representative household surveys. Since they also collect information about different sources of household income, it is possible to re-estimate national poverty rates based on incomes after taxes but before transfers. Comparing poverty rates before and after transfers gives us an idea about how effective social policies are in reducing poverty.
This is a straightforward statistical exercise, but it should be taken with a grain of salt because it does not tell us what the world would look like if all benefits were suddenly withdrawn or if they had not existed in the first place. In other words, it does not account for how policy changes would have affected individual behaviour.
Nevertheless, it is an intuitive way to illustrate the difference that transfers make to household incomes across countries and Eurostat routinely publishes at-risk-of-poverty rates before and after transfers, based on 60% of the national post-transfer median.
Subtracting the post-transfer poverty rate from the pre-transfer poverty rate gives us the absolute difference. However, to account for the fact that some countries have much higher pre-transfer rates than others, relative per cent reduction in child poverty may be more illustrative.
For instance, in 2013 child poverty rates in Greece would have been 27% (or 10 points) higher if they were calculated using pre-transfer household income. The greatest relative difference is observed in Finland, where the child poverty rate would have been 69% (or 20 points) higher. Cash transfer policies in Southern European countries as well as in newer EU member states tend to be less effective in reducing child poverty, while they are more effective in most Nordic countries as well as in Ireland and the UK.
Changes in the absolute difference that social transfers make to child poverty rates can also be tracked over time. Between 2008 and 2013 cash transfers became less effective in reducing relative child poverty in about one-third of EU member states, including countries like the Czech Republic, Hungary, France, Slovakia and Sweden. In one-third of the countries, including Germany and Norway, there were no substantial changes over time, while in another third cash transfers appear to have become more effective in reducing child poverty.
However, these differences are not necessarily due to changes in the structure or generosity of cash transfer systems, as they may also be driven by the amount of poverty to be reduced. When many people lose their jobs and start receiving unemployment or social assistance benefits, it may appear that transfers became more effective in reducing poverty even if the social protection system remained the same. For instance, it appears that cash transfers became more effective in Greece during the recession, but post-transfer child poverty rates increased from 23% to 29%, based on 60% of the median income each year.
Finally, to see if transfers became more effective for pensioners than for children, this exercise is replicated for individuals aged 65 and over. In the overwhelming majority of EU member states, social policies indeed appear to be more effective in reducing elderly poverty over time. Unsurprisingly, this is driven by pensions as opposed to other cash benefits. Interestingly, in Bulgaria, the Czech Republic, Denmark and Ireland, pensions became less effective in reducing pre-transfer poverty just as other transfers became more effective.
Thus social transfers became more successful over time in reducing poverty among the elderly in more countries than they did with regard to child poverty. In fact, a simple “difference in differences” calculation between the estimates in the above two figures shows that only in three countries – Luxembourg, Poland, and Switzerland – policies became more effective in reducing pre-transfer poverty among children than pensioners.
Disclaimer: the views expressed in this post are the responsibility of the author alone.