1
MENDING NETS IN THE SOUTH
ANTI-POVERTY POLICIES IN GREECE, ITALY, PORTUGAL AND SPAIN
Manos Matsaganis
1
, Maurizio Ferrera
2
, Luís Capucha
3
& Luis Moreno
4
1
Department of Economics, University of Crete, Rethymno GR-74100, tel: +30-28310-
77414, fax: +30-28310-77406, e-mail: manos@econ.soc.uoc.gr (corresponding author)
2
Department of Social & Political Studies, University of Pavia & Centre for Comparative
Political Research (POLEIS), Bocconi University, Milan
3
Department of Sociology (CIES), Higher Institute for Business & Labour Studies
(ISCTE), Lisbon
4
Comparative Policy & Politics Research Unit (UPC), National Research Council (CSIC),
Madrid
Abstract
The marginal role of social assistance and the absence of minimum income programmes
have long been thought to constitute defining characteristics of the southern European
model of welfare. Nevertheless, over the 1990s significant innovations in this field have
taken place. The paper aims to contribute to the analysis of recent developments by
critically examining the experience of anti-poverty policies in Greece, Italy, Portugal and
Spain. It is argued that the “patchiness” of safety nets in southern Europe is due to a
unique set of constraints, the most relevant of which are the role of families and the
“softness” of state institutions. A review of national profiles reveals that new policies
introduced in all four countries mark progress towards redressing some of the historical
imbalances of that welfare model. In particular, fully-fledged minimum income schemes
now operate in Portugal and in certain Spanish regions, while an experiment involving a
number of Italian municipalities is still in progress. In view of this, the paper concludes
that social safety nets in southern Europe remain frail in terms of institutional design as
well as political support and legitimacy.
Keywords
Poverty, social assistance, minimum incomes, southern Europe
Introduction
The reform of the welfare state, one of the most successful and resilient institutions of
last century, continues to be a bitterly contested issue. As the debate on the future of
the “European social model” gathers pace, attention is focused on pensions and other
core programmes. In contrast, social assistance – the focus of this paper – remains a
relatively neglected topic by policy makers and analysts, despite a recent emphasis on
fighting poverty and exclusion. However, the powerful forces that drive welfare reform
(namely, the slow transition from one type of labour market, family and social protection
configurations to another) also work to increase the relative importance of social
assistance within the welfare state as a whole.
The arguments are well rehearsed. The foundations of the “golden age of welfare
capitalism” in the post-war era can no longer be relied upon. In particular, the end of
“fordism” and the rise of the “new economy” have dealt a heavy blow to the labour
market underpinnings of the welfare state (Esping-Andersen 2002). Social protection
systems rested on the assumption that the (labour) market would provide steady
incomes to most workers, so that the (welfare) state could limit itself to protecting those
too old or too young to join the labour market, or those unable to do so because of
2
illness or disability. However, the generalised expectation of lifelong employment, often
with the same employer, has been replaced by rising insecurity, frequent job change,
long unemployment spells or the spread of precarious work (Taylor-Gooby 2001). Where
the bulk of social protection remains linked to occupational status, labour market
instability often translates into poverty and exclusion (Ferrera et al. 2000).
While economic change undermined the labour market foundations of the “male
breadwinner model”, social change made the domestic arrangements underpinning it
increasingly less prevalent. Higher age at marriage, fewer children per couple, increased
marital instability and the other manifestations of the “crisis of the family” have
undermined the traditional assumption of a working husband supporting a housewife and
their two or more children. Traditional families often acted as a redistributive mechanism
(pooling resources in favour of members in need), and as a provider of social services
(directing female unpaid work to the care for children, the old, the sick). As modern
families become less able (and perhaps, less willing) to perform such functions, the
pressure to formal systems of social protection intensifies (Lewis 2001). Needless to add,
demographic change compounds such pressure in the form of higher demands on
pension, health and social care systems.
The rise of atypical careers and non-standard household forms put in question the
capacity of current arrangements to support incomes and to prevent poverty. As the
occupational attachment of workers and the family attachment of dependants required
by conventional systems of social protection cease to be the norm, effective and well-
designed social safety nets become the key to a successful strategy against poverty and
exclusion. In the light of these trends, social assistance – the component of the welfare
state best suited to poverty relief – is expected to rise in prominence (Saraceno 2002).
Social assistance in southern Europe has often been described as “rudimentary” (Gough
1996, Leibfried 1993). This is not entirely unjustified: after all, social safety nets there
clearly lack the pedigree of their counterparts in northern Europe. Nevertheless, recent
developments suggest that a more complex analysis may be due.
This paper aims to contribute to such an analysis by critically examining the experience
of anti-poverty policies in Greece, Italy, Portugal and Spain. Its structure is as follows.
The next section reviews constraints to creating effective social safety nets in southern
Europe. Section three offers a brief account of policy innovations in the four countries.
The paper concludes with a discussion of unresolved issues and their implications for
future policy.
Policy constraints
The issue of strengthened social safety nets has particular resonance in southern Europe.
The marginal role of social assistance, identified as a key characteristic of south
European welfare states (Ferrera 1996, Rhodes 1996), leaves their anti-poverty armour
vulnerable. As 1996 data show, “social benefits other than pensions” reduce poverty by
a mere 1 and 3 percentage points in Greece and Italy respectively, though their effect is
stronger – but still below the European average – in Spain and Portugal.
[TABLE 1]
Poor anti-poverty performance is partly linked to limited reach of those in poverty: in
Greece and Italy, where the problem is most serious, only 31% of persons in the lowest
income quintile received “social benefits other than pensions” (Marlier & Cohen-Solal
2000). As this figure implies, many poor households are ineligible for social assistance
because they fail to fulfil the narrow categorical conditions set out by the various
programmes. Those affected include the long-term unemployed (whose eligibility to
benefit has been exhausted), new entrants to the labour market (ineligible for
unemployment insurance because never employed), the precariously employed (with no
social entitlements to draw upon in the event of temporary loss of earnings) and others.
3
Descending into poverty through holes in the safety net is a common experience among
immigrant workers and their families. Though registered foreign workers theoretically
enjoy full social rights, illegal ones have nowhere to turn except to the emergency
services provided by charitable organisations and/or by informal support networks
operated within their communities (Baldwin-Edwards & Arango 1998). Among other
factors, the “less eligibility” of immigrant workers is an inevitable effect of a social
protection regime that continues to rely on formal employment, the insurance principle
and the extended family. Seen in this light, the creation of an effective, universal safety
net assumes additional importance, as it becomes instrumental for the enfranchisement
of “outsiders” and the maintenance of social cohesion.
However, it is important to acknowledge that the “patchiness” of social safety nets in
southern Europe is no mere symptom of a more general under-development of welfare
institutions. On the contrary, the relative neglect of a comprehensive anti-poverty
dimension has often coincided with the steady growth of total social spending. If
anything, certain social programmes (for example pensions in Italy or Greece) are over-
developed, to the point of crowding-out investment in other policy areas. Therefore, it is
elsewhere that the causes of the low profile of social assistance in southern Europe must
be sought: in the unique set of constraints that inhibit its development. The two most
relevant of these are the role of the family and the “softness” of state institutions.
South European families historically functioned as an effective (though informal) safety
net: a social “shock absorber” active across a whole range of policy areas such as child
care, unemployment assistance, care for the elderly, housing or social assistance. There
is evidence that resource pooling has intensified over recent years (Fernández Cordón
1997). Nonetheless, as the family itself comes under stress, its endurance as a provider
of home-made welfare becomes uncertain (Moreno 2002). In any case, “familialism” is
known to rely on unpaid female work: often caring for children or older relatives is only
possible at the expense of erratic careers or full withdrawal from the labour market. Low
rates of female employment, especially in Spain, Italy and Greece, clearly indicate the
high social costs of the southern model of welfare (Saraceno 2000).
On the other hand, the delivery of targeted benefits requires a degree of administrative
capacity that is often simply unavailable in southern Europe. Specifically, administrative
systems suffer from low implementation capabilities, caused in part by a paucity of
resources available to street-level administrators. Moreover, the low level of political
autonomy of the administrative system in some parts of southern Europe may make it
difficult for officers in charge of benefit delivery to stand up to external pressures. As a
result, the relationship between benefit administrators and beneficiaries is in some parts
of southern Europe mediated by “brokerage” structures. To these, one must add the
poor integration of social assistance (administered by different authorities and subject to
different rules) which has created a structure allowing eligibility overlaps and gaps in
coverage.
Since social assistance benefits are typically granted on the basis of a means test, the
ability of administrators to assess “need” with some degree of accuracy is an absolute
requirement. In this sense, southern Europe presents a real challenge to social policy: a
variety of factors such as extended households, high rates of self-employment, large
informal economies and endemic tax evasion combine to create a peculiar situation. As a
result of that, administrators may be unable to judge the material circumstances of
applicants and thus their “real” eligibility to benefits (Atkinson 1998).
As the above implies, the construction of social safety nets in a context characterised by
complex socio-economic patterns, low administrative capacity and persistent tax evasion
faces specifically “southern” dilemmas (Addis 1999, Aguilar et al. 1995). As a result of
these, a straight transfer of policy know-how from more highly developed systems of
social assistance in the North would be incapable of offering satisfactory answers to such
questions. In effect, policy makers in southern Europe are left with no alternative but to
search for original solutions.
4
To a rather considerable extent, this is precisely what happened over recent years. The
renewed emphasis on anti-poverty policies led to a range of policy innovations in south
European social assistance in spite of the structural difficulties mentioned above. This
policy shift is typified by the spread of minimum income schemes, the absence of which
was thought to be a defining feature of the “southern model of welfare” (Gough 1996).
The successful launch of Revenue Minimum d’Insertion in France in 1988 set in motion
developments that led to the adoption of similar schemes throughout southern Europe.
Variations of RMI were adopted in Basque Country in 1988, in Catalonia in 1990 and in
other Spanish regions later. A national pilot scheme was introduced in Portugal in 1996
and became fully operational in 1997, while in Italy a formal experiment was started in
1998 and extended further in 2000. Only in Greece has this trend so far been resisted,
but even there the issue of strengthening the safety net is gaining in visibility. These
national policy trajectories are put in context and briefly reviewed below.
National profiles
Greece
The restoration of democracy in 1974 ushered in a period of welfare state expansion,
accelerated after the socialist landslide in the 1981 general election. The unprecedented
growth in social spending was a response to the expectations nurtured by large sections
of society over decades of politically motivated discrimination. Greece’s accession to the
European Community in 1980, widely considered to be a guarantee of political stability,
legitimised aspirations for levels of income and social protection comparable to those
enjoyed by other Europeans (Guillén & Matsaganis 2000).
Today, a high and rising level of social spending rapidly approaching the EU average is
combined with a weak performance of social transfers in terms of poverty reduction. This
apparent contradiction can be attributed to the nature of the country’s system of social
protection. Contributory social insurance is perfectly suited to “fordist” norms of long and
uninterrupted careers. By the same token, the long-term unemployed, the young who
have not yet worked, women with a patchy working history, individuals employed on
temporary or part-time basis, illegal immigrants, workers in the shadow economy and
others become “social insurance outsiders” who lose out in welfare terms, often heavily.
As a matter of fact, the welfare state in Greece places great emphasis on contributory
benefits, with little provision for non-insurable social risks such as poverty, while social
services remain at an early stage of development. As a consequence, while pensions
account for the greatest part of social transfers, policies aimed to families with children,
the disabled, the unemployed and others at risk of poverty are far less developed.
Social assistance has remained marginal. Benefits are poorly integrated, administered as
they are by different agencies and subject to different rules. Their interaction leaves in
place not a coherent whole, but an uneven structure that combines eligibility overlaps
with coverage gaps. Given that non-contributory transfers are more naturally suited to
the pursuit of anti-poverty objectives, the marginal nature of social assistance leaves a
social safety net that in reality is full of holes, through which individuals and their
families can slip into poverty. Poor households are ineligible for one of the existing
benefits if they do not fit to the “identikit” imagined by legislators, failing to fulfil the
narrow categorical conditions required.
The safety net in old age is patchy. Those with sufficient contributions are entitled to a
minimum pension plus an income-tested supplement. Lower non-contributory pensions
are paid to farmers and to those with low income and no other pension entitlement. No
universal minimum guarantee is available. Partly as a result, Greece features a unique
combination of high spending on pensions and high poverty in old age, at 13% of GDP
and 35% of those over 65 respectively (Matsaganis 2002).
5
Fragmentation and incomplete coverage are evident in all other areas of social security.
Unemployment benefit is contributory and of limited duration (12 months), as a result of
which only 44% of registered unemployed claimed benefit in 1999. Income transfers to
families are targeted to those with three children or more, so that poor children in
smaller families receive little or no assistance. Disability benefits vary by condition and
recipient status (10 categories and 22 sub-categories of benefit). Housing assistance is
contributory and geared towards owner occupation, i.e. beyond the reach of the poor.
Overall, non-contributory benefits accounted for 16.3% of all spending on social security
in 2001, while income-tested benefits for a mere 4.7%. The gradual phasing-out of basic
farmers pensions since 1998 and the abolition of the income test on “many-children
benefits” in 2002 (9.3% and 1.9% respectively of all expenditure on social security) will
further reduce the space reserved to these two types of benefits within Greece’s social
protection system.
Yet, selectivity has become a fashionable idea since 1996, when the socialist government
under a new leadership declared EMU membership an overriding aim, while pledging its
commitment to a “cohesive society”. Indeed, the concept of selectivity was hit upon as a
rather obvious way to square the circle. The strategy yielded some early results (the
income-tested pension supplement mentioned earlier), but soon ran out of steam,
presumably for lack of obvious targets in a social protection system still dominated by
contributory benefits.
The 2001 National Action Plan implicitly ruled out the option of minimum income, while
at the same time reiterating a commitment to selectivity. Three new measures, to take
effect from 2002, were announced. The most promising was unemployment assistance
for older workers, paid for 12 months to long-term unemployed aged 45-65 in low-
income families. Nevertheless, ten months after the scheme’s launch only 711 of the
35000 unemployed workers officially expected to claim had actually done so. Very low
take up beset another high-profile scheme, the social contribution rebate for minimum
wage earners introduced in 2000.
On the whole, the “danger that some groups experiencing poverty may not be eligible
for income support” (CEC 2001) remains largely undiminished despite some recent
improvements. The absence of a last resort benefit, targeted in nature but universal in
scope, remains a crucial missing link in the social safety net. Opposition to minimum
income renders the anti-poverty armour of the social protection system vulnerable.
While the administrative difficulties involved in implementing such a scheme must not be
underestimated, the financial requirement could be modest: a simulation exercise put
the cost of transfers under a minimum income programme in 2000 at €269 million or
0.23% of GDP (Matsaganis et al. 2001).
Italy
Poverty and social exclusion, marginal in the national debate and policy agenda in Italy
before the 1990s, has gained increasing salience more recently. A standard diagnosis of
the historical failings of assistenza has emerged, widely shared by political actors and
social partners: high fragmentation, policy overlaps, a bias towards transfers (against
services), marked territorial differentiation, the absence of a safety net of last resort.
In 1977 responsibility for social assistance was devolved to regional and local tiers of
government. Guiding principles and general standards were left for consideration by a
national framework law that would regulate social assistance, which was not issued until
2000. Laws passed by various regions allowed wide discretion at municipal level. Local
minimum income schemes were the product of municipal initiative. Turin in 1978,
Ancona in 1981, Catania in 1983 or Milan in 1989 introduced a non-categorical means-
tested benefit known as minimo vitale, even though many other municipalities (such as
Bari or Rome) did not.