Opinion: Stress-testing welfare states

Stress testing welfare systems by modelling actual or hypothetical shocks to income can help assess individual vulnerability to poverty while helping to reduce the overall risks to society.

Measuring vulnerability: A need for forward looking analysis

The full extent of individual vulnerability is revealed when common economic shocks occur. The latest global economic crisis was an extreme example of a stress test to many welfare states. To give a vivid example, the emigration flows stripped my home country Lithuania of around 7% of its population between 2009 and 2012 – the period when the crisis hit hardest.

As more social statistics, survey data and research on the consequences of the crisis emerge, the devastating effects have become apparent. That, however, is a backwards looking analysis. What is needed is forward-looking analysis of the ability of the welfare state to mitigate acute shocks to individual and household incomes.

Anthony B. Atkinson (2009) has proposed just that. Borrowing from the financial sector, he suggests using the method of stress-testing. Adopted properly, stress-testing can monitor changes in social protection policies and related shifts in individual vulnerability to poverty. This will also help mitigate general social risks and help ensure shared prosperity.

So how does it work? In brief, the approach utilizes tax-benefit micro-simulation techniques to model actual or hypothetical shocks to income. The shocks can be simulated on the household survey data representative of the country’s population or, if available, administrative micro data. It enables the examination of the likely impact of income loss on living standards. It takes into account interactions between individual and household characteristics, and the design of the tax-benefit policy instruments.

The approach has so far been used to analyze welfare compensation for unemployment, modelling the impact of the recent recession (Figari et al. 2011; Fernandez Salgado et al. 2013). Others have looked at the effects of tax and benefit systems on the bottom 40% of low income earners (Frölich, Kaplan, Pagés, Rigolini and Robalino 2014).

Stress testing should not be restricted to modelling actual shocks and their effects. Hypothetical income shocks can be simulated for wide population groups, by focusing on the most recent or even planned changes in the tax-benefit rules. Income loss related to unemployment, childbirth, family dissolution, involuntary reduction of working hours, disability, retirement and other risks may potentially be analyzed.

Stress-testing welfare states in the EU – a lesson not to be missed

The good news is that there is a tax-benefit micro-simulation model EUROMOD that is currently freely available for the 27 EU member states. Estimating people’s vulnerability to poverty measures using the model’s component for Lithuania gave promising results (Navicke 2014). The research looked at two events which can have a dramatic impact on family incomes – unemployment, and the loss of the mother’s income upon childbirth. The estimates highlighted stark inconsistences in the levels of protection provided against different risks in Lithuania.

he analysis also revealed how both levels and profiles of vulnerability related to the design of the tax-benefit instruments directly and indirectly targeted towards mitigating the analyzed risks.
To illustrate, one in three households at the mid-point of Lithuanian income distribution proved to be vulnerable to falling straight into poverty in the case of the unemployment of one of their insured members. The probability of poverty risk was around 50% for the sole earners that otherwise took a median position in society. On the other hand, in the case of childbirth, this probability was close to zero.

Moreover, the social protection system was so generous as to lift a fraction of those with newly born children above the poverty threshold. While there is no doubt that investment into newborns is important, children do stay in the families for longer than just the initial few years.Therefore, a better balanced social protection system could help prevent parents from emigrating once losing jobs later in life.

Further research is underway. A comparative cross-country analysis of vulnerability profiles across the EU is foreseen in the framework of the Norface project scheduled for 2015-2017. These new tools and measures of vulnerability to poverty will provide further insights into the functioning of the different welfare states, as well as the impacts on individual wellbeing and economic prosperity.

—-

Jekaterina Navicke is on the Department of Sociology at the University of Vilnius and a member of the EUROMOD consortium.

The Wold Bank’s Jobs and Development Blog