Are levels of social welfare payment the problem or is it something else?

Posted on May 13, 2017 by Tom Healy

Tom Healy, Director NERI
Tom Healy, Director NERI

‘Why don’t they get a job?’  “Isn’t social welfare making people chose ‘leisure’ over ‘work’?” ‘Aren’t social payments in Ireland way above those in the UK and elsewhere?. Only a particular perspective and frame of mind among economists, politicians or commentators trigger these sorts of questions. The mentality is by no means confined to those who enjoy a comfortable lifestyle and degree of economic security. Even in areas of acute social deprivation, there can be a culture of resentment, suspicion and prejudice. This may take the form of prejudice on racial, age, gender or marital status. 

Witness the commonly held view that a young male dressed in a particular way and living in a particular locality is prone to idleness and ill-gotten activity (or lack of any activity). Witness the commonly held view that single mothers do well out of social welfare and chose a life of dependence on the State.  And, then, the view that migrants come here to cash in on ‘our’ social welfare and other entitlements. Social prejudice is born out of fear, ignorance and sweeping generalisations. It is in this context that debates about the ‘incentive to work’ are conducted in the rarefied context of economic analysis. But, what do the statistical facts show?

A useful, but limited, concept is the Net Replacement Rate. It is defined with reference to the level of income of individuals or households that would be received following the loss of income due to unemployment. It is calculated as a proportion of previous after-tax income and may or may not include housing cash benefits (such as rent supplement in the Irish case). To make things simple, an arbitrary measurement of 67% of average (median) earnings from employment is often used as a benchmark. Assuming that the NRR is lower than the benchmark, the higher the NRR is, the smaller the difference between what people would receive – on average – before and after losing a job. Typically, comparisons are done for specified household types (single, married, children etc.). In some cases, the NRR might be calculated as being higher than the average wage in which case for that household or individual type there is no ‘incentive to work’ on this narrow measure.

In practice, matters are more complicated. For one thing, people look for and take up jobs for many reasons. Cash income is a huge and primary factor. However, the complications of living are many and encompass caring roles, sickness within families, access to childcare, transport and other social supports. Moreover, loss of income – whether short-term or long-term – may be accompanied by eligibility for support in areas such as healthcare, housing and education. Taking a full and realistic account of all of these factors is important but difficult especially in an international comparative setting. The striking aspect of the matter is that even on a narrow measure of ‘incentive’ based on NRRs the Republic of Ireland and the United Kingdom stand out as having low rates compared to most other European countries.

The saying that comparisons are odious applies here.  Depending on which family type (single, couple, no children, with children), what benchmark is used (% of average earnings as the reference before loss of job) and what benefits are including (housing or not), which time periods are chosen (immediately or 5 years after loss of job), which years are selected and which selection of countries (all EU28 including Bulgaria and Estonia, advanced EU states only, OECD including Chile and Turkey or UK-Ireland only). Take your pick!

If we take data published by OECD for 2014 and focus on net replacement rates during the initial phase of unemployment and apply different assumptions in regards to the wage benchmark some interesting patterns emerge:

(source OECD Benefits and Wages statistics – “During the initial phase of unemployment” )

What do crude data tell us? In Table 1 below, I have ranked thirty-six household types and various average wage benchmarks by the value of the Net Replacement Rate in the Republic of Ireland during the initial phase of unemployment. In all, 36 cases are shown covering different combinations of assumptions with regards to family type, wage benchmark and housing assistance. Out of 36 cases in 2014 (the latest year for which OECD data were available), only one household type showed an NRR in excess of 100: a one-earner married couple without children and benchmarked on 67% of the average wage. Only five household types showed a higher NRR than 80%. These all comprised married couples.  The typical ‘villains’ of social welfare, the lone parents and single persons, were nowhere to be found at the top of the ranking. Rather, they were concentrated overwhelmingly at the bottom. Stereotypes must work harder to prove someone's point!  What about NRR values based on long-term unemployment benefit?  Here, an interesting pattern emerges. The NRR values and household ranking is almost identical between initial and long-term scenarios. In other words, the hit to income is sharpest for Irish household in the initial phase of unemployment but remains fairly constant thereafter. Even still, lone parents and single persons show the lowest NRRs and with one household type (couples without children and in receipt of housing assistance) have an NRR value in excess of 100. USE THE SCROLL BAR IN THE EMBEDDED TABLES BELOW TO SEE ALL OF THE DATA. THE DATA ARE RANKED BY SIZE OF ESTIMATED NRR.

A very different pattern exists in the European Union.

First of all, it is useful to rank household types of the ‘ratio of relative generosity’ which I define as the ratio of Irish NRR to EU (median) NRR. I start with the initial phase of unemployment comparison (Table 2).  Of 36 household types, 27 Irish household types show lower NRR values compared to the EU median.  In other words, during the initial phase of unemployment, most Irish households take a big hit and with the exception of married couples with one earner, experience lower replacement rates than the EU comparators.

Second, of all, I show the ‘ratie of relative generosity’ for households based on long-term benefits . A clear distinction opens up between Ireland and the EU average. In all cases for which data were available, the ratios of NRRs are above 100. In other words, though Irish households would be worse off not in employment the ratio of benefits to earnings is relatively favourable in the case of Ireland (table 3).

Does this explain patterns of long-term unemployment or what economists call ‘economic inactivity’? Is the solution to lower and restrict benefits to encourage or incentivise long-term unemployed to seek work? Space does not permit a thorough investigation of these questions. However, we can be confident that any programme of systematic benefit cut-back would expose all households, but some in particular, to additional poverty and deprivation. If there is a problem with NRRs in the long-run then the answer is on the wages side and not welfare.  And we need to add in ‘social wages’ (loosely defined as those collective goods such as housing, transport, health and education which are provided socially).

However, the capacity of some other EU member states to absorb unemployment through relatively generous income support in the immediate months or year following loss of employment combined with systems of support, training and provision to enable rather than punish the long-term unemployed seems to provide a better approach to encouraging high employment rates and low poverty rates among all households types. The ‘stereotypes’ have to look to better examples of better designed public policies.

Digital Revolutionaries