SLIDE 1
Social safety nets in (the) crisis
Anti-poverty policy in Greece 2010-2015
Manos Matsaganis Associate Professor Athens University of Economics and Business CSP Lunch Seminar University of Antwerp 29 January 2016
SLIDE 2 The Greek crisis and austerity in brief
Greece entered into a bailout agreement with the EU, the ECB and the IMF (May 2010), in return for massive spending cuts and tax hikes, and a programme of structural reforms.
The bailout agreement has (so far) enabled Greece to remain part of the Euro area – at the cost of a severe recession, widespread misery, and political instability.
Other ‘programme countries’ such as Latvia (January 2012), Ireland (December 2013) and Portugal (June 2014) have now successfully exited their programme.
But Greece’s attempts at an early exit (October 2014) under the previous conservative-socialist coalition government, and renegotiation (February- July 2015) under the new radical left-nationalist right coalition government backfired badly. A new three-year programme was signed in August 2015.
SLIDE 3
economic effects GDP contraction
In recent years the Greek economy shrank enormously: in 2014 GDP had fallen 26.0% below its 2007 level.
This only compares with the US Great Depression (approx. -30% in 1929-1933).
In other countries badly hit by the crisis, GDP fell less and/or recovered faster (-6.8% in Portugal, -7.5% in Latvia in 2007-2014).
Gross domestic product at market prices, in real terms (2007=100). Source: Eurostat.
SLIDE 4
economic effects divergence
In 2000 GDP per head in Greece was at 74% of the EU-15 average.
At the pre-crisis peak (in 2009) relative income had converged to 86%.
By 2015 Greece had fallen back to 64% of the West European average.
Gross domestic product at current market prices per head of population (PPS: EU-15 = 100). Source: AMECO Eurostat.
SLIDE 5
labour market effects employment
Employment had always been below the EU average. But in recent years it fell precipitously: from 61.4% in 2008 to as low as 48.8% 2013.
This, in effect, undid the progress of the previous two decades and more (the employment rate in 1992 had been 53.7%).
Employment rate (%). Source: Eurostat.
SLIDE 6
labour market effects unemployment
GDP contraction caused unemployment to rise, same as everywhere. Except that in Greece it rose more sharply, peaked later and is now the highest in the EU.
In September 2015, seasonally adjusted unemployment stood at 24.6%.
Unemployed rate (%). Source: Eurostat.
SLIDE 7 labour market effects earnings
Real wages also fell, both as a result of the recession (reduced demand for labour) …
… and in the context of ‘internal devaluation’ (policy-driven compression of wages via labour market deregulation).
The decline in gross earnings in 2008-2013 was 24.6% in Greece (vs. 12.3% in Portugal).
- Source: Myant & Piasna (2014) ‘Why have some countries become more unemployed than
- thers?’ ETUI Working Paper 7 2014.
SLIDE 8
labour market effects earnings growth by category
In Greece, earnings growth in 2000-2009 was undone by earnings decline in 2010-2014. On average, earnings in 2014 were 8.4% below their 2000 level.
Trends differed by category: workers in public utilities had been awarded fabulous pay rises in the past (and still remained 3.4% above the 2000 level).
Source: Bank of Greece.
SLIDE 9 labour market effects internal devaluation
Recent labour market reforms in Greece included a legislated cut in the minimum wage (February 2012):
(-32% for young workers)
Also, sweeping deregulation of labour market institutions:
collective bargaining
employment protection legislation
etc.
Changes especially affected those newly hired and/or young workers:
administrative data from IKA (the social insurance agency for private sector employees) indicate that in 2010-2014 median reported earnings of all insured workers fell by 18.3% in real terms.
real median earnings were cut less (by -9.6%) for those who kept their job with the same employer.
for those aged below 30 the decline was -34.8% (newly-hired workers: -34.2%).
SLIDE 10 poverty effects three indicators
We estimate poverty effects in 2011-2015, using EUROMOD, in terms of three different indicators.
The standard poverty rate shows the proportion of the population with a net equivalent disposable income below 60% of median.
- 2015 threshold: €824 per month for a couple with two children
The second indicator ‘anchors’ the poverty line at 60% of the 2009 median, adjusted for inflation.
- 2015 threshold: €1,271 per month for a couple with two children
The third indicator measures the proportion of population who are unable to purchase the cheapest basket of goods consistent with dignified living without dissaving, borrowing, or getting into debt.
- 2015 threshold: €640 per month for a couple with two children living in Athens, in owner-occupied
housing
Source: PARU estimates.
SLIDE 11 poverty effects three indicators
Source: PARU estimates.
In 2015, our estimated headcount rate was:
- 23.2% (relative poverty)
- 45.3% (anchored threshold)
- 15.2% (extreme poverty)
SLIDE 12 poverty effects headcount rates
Source: PARU estimates.
poverty rate 2015
population share
relative anchored extreme all households
100.0%
23.2% 45.3% 15.2% gender men
49.1%
23.4% 45.3% 15.8% women
50.9%
23.0% 45.2% 14.5% age 0-15
15.4%
23.6% 46.2% 17.4% 16-29
15.6%
33.9% 57.1% 24.2% 30-44
23.0%
22.8% 42.7% 16.5% 45-64
26.5%
25.5% 46.2% 16.5% 65+
19.5%
11.6% 36.8% 2.7% household head is unemployed
8.3%
80.6% 91.0% 70.4% employee (public sector or banking)
11.8%
1.5% 10.2% 0.7% employee (private sector excl. banking)
23.7%
17.7% 45.4% 12.2% liberal profession
3.3%
7.8% 13.2% 4.9%
10.6%
32.1% 54.3% 23.8% farmer
6.7%
38.7% 66.3% 18.3% pensioner
32.9%
13.3% 39.6% 3.8% inactive, student, other
2.6%
55.3% 77.6% 43.7%
SLIDE 13
poverty effects focus on extreme poverty
We focus on extreme poverty.
Arguably more relevant for policy (?)
The extreme poverty threshold converged to 50% of median income in 2013 (47% in 2015).
This in spite of the fact that our basket of goods is very Spartan.
Note that the corresponding estimate for food for a healthy diet & kitchen equipment alone under the ‘EU reference budgets’ project was €915 per month (67% of median).
Source: PARU estimates.
SLIDE 14
poverty effects extreme poverty
Somewhat surprisingly, we find that extreme poverty rates were much lower for the elderly (2.7% in 2015) than for children (17.4%), while they were alarmingly high for households with unemployed head (70.4% in 2015).
Source: PARU estimates.
SLIDE 15
In principle, a well-designed system of social protection should be able to mitigate the social effects of an economic crisis.
“These are precisely the kinds of emergencies that welfare state programmes and institutions are designed to deal with, so that when a financial crisis turns up we have routine mechanisms […] for coping with its consequences. Long lines of the unemployed caused by economic crises are the core business of the welfare state”
- Source: Castles F.G. (2010) ‘Black swans and elephants on the move: the impact of
emergencies on the welfare state’. Journal of European Social Policy.
But in Greece, as the crisis increased the demand for social protection, the austerity reduced its supply.
Or did it?
policy responses the core business of the welfare state?
SLIDE 16 policy responses social expenditure (% GDP)
Source: Eurostat.
In terms of social expenditure as % of GDP, things look rather encouraging.
Greece now spends more than the EU-15 (and much more than Portugal).
- 31.6% vs. 29.5% respectively in 2012 (Portugal: 27.6% in 2013)
- Latest figures out of date!
SLIDE 17
policy responses social expenditure (per head)
Source: Eurostat.
But in terms of euros spent per inhabitant (in real terms, 2005 prices), social expenditure in Greece remains well below the EU-15 figure.
What is more: social spending is actually falling since 2009 (which is when the crisis hit the country).
SLIDE 18
So, what is going on?
Pre-crisis: rising social expenditure …
faster than GDP, and faster than the EU average
… but disappointing social effectiveness
gross inefficiencies in core programmes (pensions and health) remained
income support and social care to the poor and other vulnerable groups neglected
serious gaps in the social safety net
When the economic crisis hit the country, the welfare state failed to rise to the occasion
policy responses Greek welfare unfit for the crisis?
SLIDE 19
policy responses anti-poverty effectiveness (pre-crisis)
SLIDE 20 policy responses retrenchment and (some) expansion (2010-2014)
On the whole, the Greek programme entailed massive fiscal consolidation, especially in the context of the 2013-2014 Spending Review:
5% of GDP in 2013 and 2.25% in 2014
Social protection was identified as a key source of savings:
cuts in social spending: 45% of total savings increases in social contributions: another 5%
Some (limited) scope was left for policies to strengthen the social safety net
child benefits: comprehensive means-tested scheme introduced in 2013 unemployment assistance: eligibility conditions broadened in 2012/2014 minimum income pilot: launched in November 2014 social dividend: paid as a lump sum in 2014
Ratio of contractionary vs. expansionary measures in fiscal terms: approx. 4:1
SLIDE 21
policy responses pension spending (per head)
Expenditure on pensions (per head of population, in 2005 prices) has continued to rise.
Source: Eurostat
SLIDE 22
policy responses pension spending (% GDP)
As a share of GDP, expenditure on pensions in Greece has grown enormously.
Sustainability concerns?
Crowding-out effects (vis-à-vis other social benefits)?
Source: Eurostat; Ministry of Labour
SLIDE 23 policy responses pension benefit cuts (2010-2014)
Yet, pensions were cut significantly under austerity.
Higher pensions hit more …
… even though low-income pensioners also affected.
Source: Ministry of Labour; own calculations
level of monthly pension (in 2010) annual pension income in real terms (€) 2010 2014 change (%) €600 pcm 8,400 7,069
€800 pcm 11,200 9,079
€1.300 pcm 18,200 13,898
€2.200 pcm 30,800 19,611
€2.500 pcm (age >65) 35,000 20,500
€2.500 pcm (age <55) 35,000 19,059
SLIDE 24 policy responses effective retirement age (March-May 2015)
Nevertheless, the number of pensioners kept growing (to 2.7m)
(as the number of workers continued to shrink to 3.6m)
… while retiring at a tender age continued to be easy
25% of all retirees in spring 2015 were aged below 55 (in the civil service: 34%)
Source: Ministry of Labour; own calculations
age at retirement all retirees civil servants no. % no. % <25 1,080 4% 219 4% 26-50 1,935 7% 532 9% 51-55 4,108 14% 1,319 22% 56-61 8,338 29% 2,135 36% 62-67 5,789 20% 912 15% 68-73 3,092 11% 292 5% >73 4,038 14% 601 10% unknown 16 0% 0% total 28,396 100% 6,010 100%
SLIDE 25 policy responses unemployment benefit receipt (2010-2015)
Meanwhile, other social protection programmes fared less well.
While the number of unemployed workers sky-rocketed, the number of recipients
- f unemployment insurance benefit (the main income support scheme in the
event of job loss) actually fell.
Source: ElStat and ΟΑΕΔ (Public Employment Service).
SLIDE 26
policy responses unemployment benefit coverage rate (2010-2015)
As a result, the coverage rate of unemployment insurance benefit fell to 9.5% in 2015.
The coverage rate was 11.8% if two minor unemployment benefits (for the long-term unemployed, and those formerly self-employed) are also included.
Source: ElStat and ΟΑΕΔ (Public Employment Service).
SLIDE 27
policy responses single child benefit
Until recently, the majority of families received little or no support, even when they lived in poverty. In contrast, child benefits for larger families, and family allowances for core workers, were rather generous. From January 2013, a non- categorical single child benefit has been introduced on a means-tested basis in place of large family benefits and child tax credits.
Unlike most of the programmes it replaced, single child benefit is available from the first child, and for a larger age group (up to 18; or up to 23 if in college; or irrespective of age if disabled).
The new benefit, even though welcome, is quite modest: it pays up to €40 per month to families with one child, raised to €245 per month for families with three or more children (NB: since 2012, the minimum wage is set at €586 pcm).
The number of recipients is approx. 750,000 families (i.e. 53% of all households with at least one child under 25 years).
Spending on the scheme reached 0.33% of GDP in 2013.
SLIDE 28
policy responses social dividend
In 2014, after six consecutive years in recession, the Greek economy registered a small positive growth (GDP up by +0.7%).
In view of that, the then (conservative-socialist coalition) government decided to pay a one-off lump-sum ‘social dividend’ to low-income groups to (partially) compensate them for their losses during the crisis and austerity.
The social dividend was worth €500 for a single person or €833 for a couple with two children (€647 per receiving household on average).
The number of recipients was over 690,000 households (16% of all households).
Spending on the scheme amounted to 0.25% of GDP in 2014.
In 2015 the scheme was discontinued.
SLIDE 29 policy responses dealing with the ‘humanitarian crisis’ (2015-)
The radical left-nationalist right coalition that emerged victorious from the 2015 general election(s) had made a lot of political capital by railing against the austerity-engineered ‘humanitarian crisis’.
But those who might have been led to expect that strengthening social safety nets would now be priority were in for disappointment.
The new government’s anti-poverty measures were less ambitious even relative to those introduced under the previous government.
energy subsidy (max. €40 pcm)
rent benefit (worth €130 pcm for a couple with two kids, paid directly to landlords)
food card (voucher to purchase food, worth €130 pcm for a couple with two kids)
SLIDE 30 policy responses dealing with the ‘humanitarian crisis’ (cont’d)
The total resources set aside for the government’s three measures to deal with the ‘humanitarian crisis’ amounted to €200m in 2015 (0.11% of GDP).
Actual cost turned out to be just below €130m (0.07% of GDP), which includes the projected cost of 2015 entitlements paid in 2016.
Food card accounted for the lion’s share (€114m).
The previous government’s discretionary measures – which had been rightly criticized as inadequate – had cost more:
- €654m (0.36% of GDP) in 2013
- €1,135m (0.64% of GDP) in 2014
Moreover:
No action to improve the coverage of unemployment benefits.
No enthusiasm for continuing the guaranteed minimum income pilot.
SLIDE 31
Income losses might have been less painful if the victims of the recession could at least rely on the public provision of health and other vital social services (free of charge – or nearly so - at the point of use).
Given the extend of waste and inefficiency before the crisis, it would not have been impossible to cut costs while at the same time maintaining an adequate quantity and quality of service.
As it turned out, the opposite has happened:
austerity cuts have been largely indiscriminate
reforms have been disruptive and/or have raised co-payments
most suppliers have reacted to arrears by withholding supplies
industrial action has contributed to the general unreliability of public provision
some public sector workers have reacted to wage cuts by reducing effort (though most others work more for less)
policy responses publicly-provided services
SLIDE 32 policy responses local provision
Local initiatives flourished, partly compensating for central government inaction. School meals
In response to reported cases of undernourished children, the Athens Municipal Nursery started to deliver school meals on a daily basis.
In 2012-2013, the number of children receiving school meals amounted to over 8% of all children attending the schools concerned, or approx. 2% of the total school population in the City of Athens.
Other local councils run a similar though less ambitious service. Soup kitchens
Emergency food aid is also provided in soup kitchens run by local councils, the church, and voluntary organizations.
In 2014, a survey of 500 persons attending soup kitchens in Athens found that 54% received food for themselves as well as their family, while 37% reported having children.
About two thirds of all respondents were actually Greek.
SLIDE 33 Source: Beckerman W. (1979) ‘Impact of income-maintenance payments on poverty in Britain, 1975’. Economic Journal.
policy responses evaluating anti-poverty policies
We estimate the anti-poverty effectiveness of three ‘flagship’ anti poverty policies: single child benefit (2013), social dividend (2014), food card (2015)
in terms of Beckerman’s poverty gap efficiency
using EUROMOD
SLIDE 34 Source: PARU estimates. Food card evaluated as income support, assuming perfect targeting.
policy responses estimating poverty reduction
We first measure reduction in numbers in extreme poverty, using EUROMOD.
child benefit social dividend food card all households 109,985 120,846 21,389 gender men 59,410 62,826 10,886 women 50,575 58,020 10,503 age 0-15 42,667 29,651 6,212 16-29 19,601 21,193 4,929 30-44 35,348 33,235 5,572 45-64 14,504 29,563 4,393 65+ 7,204 283 household head is unemployed 8,325 28,339 116 employee (public sector or banking) 5,851 679 employee (private sector excl. banking) 36,808 34,062 1,102 liberal profession
32,048 14,684 16,159 farmer 26,225 21,611 1,391 pensioner 20,335 inactive, student, other 4,782 882
- no. taken out of extreme poverty
SLIDE 35 policy responses estimating anti-poverty effectiveness
We then measure effectiveness in reducing extreme poverty in terms of Beckerman’s poverty gap efficiency, using EUROMOD.
child benefit social dividend food card all households 9.5% 9.1% 8.7% gender men 9.7% 9.2% 8.8% women 9.2% 8.9% 8.7% age 0-15 20.5% 10.6% 13.2% 16-29 6.8% 7.7% 9.5% 30-44 12.3% 10.4% 8.2% 45-64 4.1% 8.2% 6.8% 65+ 0.0% 14.4% 4.5% household head is unemployed 5.4% 6.8% 10.1% employee (public sector or banking) 14.1% 14.2% 0.0% employee (private sector excl. banking) 13.8% 11.4% 8.9% liberal profession 27.2% 5.3% 0.0%
17.0% 11.5% 9.9% farmer 20.2% 16.3% 3.6% pensioner 1.7% 15.4% 3.3% inactive, student, other 6.6% 8.1% 3.3% reduction in total initial poverty gap
Source: PARU estimates. Food card evaluated as income support, assuming perfect targeting.
SLIDE 36 policy responses summing up
How did the three ‘flagship’ anti-poverty policies fare? Single child benefit (2013)
extreme poverty rate reduced by 1.0 pp.; extreme child poverty rate reduced by 2.5 pp.
poverty gap efficiency: -9.5% (overall); -20.5% (children)
cost: 0.33% of GDP
net cost (relative to previous array of family benefits and tax relief): negligible Social dividend (2014)
extreme poverty rate reduced by 1.1 pp.
poverty gap efficiency: -9.1%
cost: 0.25% of GDP Food card (2015)
extreme poverty rate reduced by 0.2 pp.
poverty gap efficiency: -8.7%
cost: 0.07% of GDP
SLIDE 37 concluding remarks
Anti-poverty policy in Greece remains extremely low profile - which is rather impressive for a country having just experienced such a dramatic increase in poverty.
Under austerity, total social spending has fallen – except for pensions, which dominate the Greek welfare state even more than they have always done.
Policies to strengthen the social safety net were introduced (sometimes at the insistence of the country’s lenders), but they were mostly ‘too little too late’.
Large protection gaps still evident
- too few unemployed workers receive income support (even when they are very poor)
- Greece still lacks a social assistance scheme of last resort (GMI-type or equivalent).
SLIDE 38 concluding remarks (cont’d)
General indifference towards a more effective anti-poverty policy broadly shared across the political spectrum
- the socialists neglected poverty for far too long
- the radical left has used the rhetoric of impoverishment to justify the status quo
Political economy of reform
- the poor not influential: voiceless / disenfranchised
- pensioners / recipients of disability benefits: more powerful lobbies
- farmers / providers of public services: virtual veto points
- (middle-class groups: medical / legal / engineering professions)
- governments like to take the path of least resistance
- (Greek ones apparently more so …)
