The social security number: A small device underpinning ...

The social security number:A small device underpinning

big systemsPaul-Anthelme Adèle

University of Nantes, France

Abstract The possession and use of a personal social securitynumber helps to structure people’s daily lives. However,despite its fundamental normative importance, the socialsecurity number remains a little-known entity. Increasinglyuniversal and yet diverse in form, it is a legal and technicalnorm which is as much a mechanism for surveillance andmonitoring as it is a necessary instrument for giving effect tosocial rights. Analysis of this constituent element of socialsecurity systems permits as assessment of some of thetechnical difficulties presented by the ever-increasingmovement of people and data. Overcoming these technicaldifficulties should permit to envisage a first technical steptowards realizing a universal and global social security system.

Keywords social security administration, legal aspect,information technology, data processing, international

Introduction

What could be more commonplace than being assigned a number? Setting upa telephone contract, booking a seat on a train or taking a ticket for a queueall figure among the countless daily acts that necessitate a number. Yet each timea person perform these acts, he or she becomes involved with a legal andtechnical process for which the normative rules are not easily discernable. Thisapplies equally to the use of a social security number. Structured by a whole

Address for correspondence: Paul-Anthelme Adèle, Maître de conférences, Université de Nantes,Laboratoire Droit et changement social (CNRS – UMR 6297), 1 Quai de Tourville, 44035 NantesCedex 1, France; email: [email protected].

International Social Security Review, Vol. 70, 1/2017

© 2017 International Social Security Association

3

series of – mandatory – legal texts, it is without doubt a legal norm. But does thatmake it a norm like any other? The answer is no; far from it.

The social security number has been addressed by the legal literature, but onlyfrom the angle of personal data protection (Council of Europe, 1986). Giventhat such a number recalls or gives access to personal data, its storage andcommunication are sensitive operations. For this reason, the creation and use ofsocial security numbers are strictly governed by law. However, this may not bethe most innovative way of looking at personal social security numbers.

First, the social security number is an element of our identity. Each individual isaccorded a unique number, and vice versa. It is necessarily linked in one way oranother to different pieces of information mentioned in our civil status, such asour sex, date and place of birth, or home address. Depending on the socialsecurity system, it might also be linked to numerous other pieces of personalinformation relating to our work, family situation or medical care.

However, as well as relaying a number of private details, the social securitynumber is an instrument that enables people to assert their rights. It is a legalnorm that certifies the material facts of our existence, which are recognized asgiving us rights.1 It belongs to those atypical norms that have a tendency tomuddy the waters around the theoretically rigorous distinction between practiceand law, between what “is” and what “ought to be”.2

Above all, the social security number is a fundamental legal device insofar as itunderlies the institution of social security as an indispensable cog in the machine.Without it, social security would be practically ineffectual in view of thesheer number of people it has a duty to protect. Nevertheless, the social securitynumber is shaped by texts that are little studied owing to their subordinate placewithin the hierarchy of norms.

Lastly, the social security number is a feature that is common to all socialsecurity systems. Analysing it therefore enables us to develop both a generaltheory of social security and a valuable comparison of national legal systems.Specifically, this analysis may be informed by national examples (e.g. Canada,France, the United Kingdom and the United States) as well as internationalexamples (European Union, International Labour Organization).

The legal view of the social security number is generally focused on membershipof a social security “system”. The allocation of a social security number and theprocess of registration are two interconnected notions which seem, at first sight,to mesh together into one legal mechanism. They should, however, be carefully

1. Strictly speaking, however, it is only the certification of these facts by a legal act that enables theinsured person to invoke them as rights to social security. These facts do not themselves generate anyrights unless they are recognized by a legally-authorized authority.2. On this logically very sound distinction between the “is” and the “ought”, the Sein and the Sollen, seeKelsen (1962); Goyard-Fabre (1986); and Béchillon (1997).

The social security number



4

differentiated. While “registration” refers to the system for identifying oneindividual within the community of insured individuals, the social securitynumber is the technical element of this system, the part that ensures that aparticular number, and that number alone, is assigned to a particular individual,and that individual only.

Behind the unity of the social security number, therefore, lie two significantintellectual operations. The first corresponds to the composition of the number:simply put, the technique for creating or devising it. The number can be viewedfrom this perspective as a minor technical device. The second intellectualoperation associated with the social security number is registration. In this light,it emerges as a legal norm of exceptional power.

A small device

Although social security identifiers vary in format, certain constants help us to seehow these legal devices function. While the social security “number” is always arather short chain of characters – around ten in most systems – this simple andlimited sequence is a powerful device for counting and identifying millions ofinsured persons.

For a better understanding of this device, it is worth briefly retracing itstechnical history to a time when information technology did not yet exist, at leastnot in its current form.

From digits to digital

Why does each personal social security identity take the form of a “number”? Afterall, each person has a civil status profile which enables her or him to be identifiedwithin a population. Surname, first name, and date and place of birth are pieces ofinformation that many national administrations held before modern social securityinstitutions were established. However, the number is in fact vested with technicalpower that far exceeds that of traditional civil status data.

First, within a social security system, limiting and standardizing the number ofprinted characters enabled considerable material savings to be made giventhe millions of individuals involved. Such savings were welcome in view of thelimited technical means of social security administrations across much ofthe twentieth century. Typically, data processing devices (for example, unitrecording equipment), the precursors of computer technology, were the onlyavailable tools. The information entered into these machines was inscribed onpieces of card known as “punch cards”.3 On each card was printed a grid of

3. See, for example, Griffe (2006); Hurtaut (1999); and SEMA (1969).




5

numbers which could, as the name suggests, be punched with holes to makecomplex combinations. Each combination corresponded to a character, numberor letter and each card could contain only a limited number of characters.4 As aresult, a system that enabled the size of the social security identifier to bereduced in a simple manner made it possible, in material terms, to create andstore millions of punch cards.

Situating the use of the available technology in a historical context brings us toanother aspect of modern-day identifiers. The term “number” can be slightlymisleading because it does not always designate a number composed of Arabicnumerals like the French numéro de sécurité sociale. The United Kingdom’s“National Insurance number” is one such example.5 Consequently, although theidentifier is systematically a sequence of characters, its general format is notuniversal.

Letters have the advantage over digits of offering greater space-saving, becausethey come from an alphabetic writing system (for instance, the 26 letters (A to Z)of the English alphabet used in this article) rather than a decimal system (0 to 9).However, the use of letters presents two disadvantages. The first is the risk ofconfusion between certain letters and certain digits; the graphical proximity ofzero and the letter “O” provides the best illustration. This obstacle may beovercome either by excluding certain letters in order to reduce the potential formisreading or by using certain typographical designs, such as the “slashed zero”.In this way, the United Kingdom’s National Insurance number disallows certainletters or combinations of letters that risk causing administrative confusion.6

The second disadvantage of letters is the greater amount of data that is requiredto “translate” a letter onto a punch card uniquely composed of columns of figuresfrom 0 to 9. While a number can be written by merely punching one of theboxes – the one corresponding to that number – writing a letter requires the useof a numerical language which translates one letter into a combination of figures.

In 2017, such concern to save technological material resources – “hardware”, ifyou like – might seem trivial.7 But it is not when it is considered from theperspective of the number of IT operations carried out every day on every socialsecurity number. To process a medical care, pension payment or social securitydeclaration, the numbers are routed, from a swipe card to a computer terminal,

4. The most common format up until the advent of contemporary information technology was a gridof 80 columns that limited to 80 the number of characters that could be expressed through punchedholes on each card.5. The French numéro de securité sociale is composed of thirteen digits divided into seven groups (e.g.1 79 08 75 114 176 92); the United Kingdom’s National Insurance number is composed of ninecharacters – two letters, six figures and a letter (e.g. AB123456C).6. For an explanation, see <www.hmrc.gov.uk/manuals/nimmanual/nim39110.htm>.7. Except in the case of the considerable concern surrounding the global energy consumption ofcomputers and networks.




6

www.hmrc.gov.uk/manuals/nimmanual/nim39110.htm

from server to server. They are encrypted, decrypted and re-encrypted multipletimes and in multiple places.

These actions could be viewed as extra-legal, merely “technical”, the stuff ofengineering discussions. But this is to overlook the fact that the social securitynumber is first and foremost a legal operation, shaped by legal norms to facilitatethe functioning of the social security system. As currently used, it assumes anelectronic format in order to extend its normative function by means of a realalgorithmic identity.

An algorithmic identity

Why use one sequence of characters and not another? In all cases, the number iscomposed using an algorithm and a systematic method. This formula offersimportant advantages, including – yet again – the considerable savings it allowsin the means required to generate and use an individual identifier. An algorithmis a series of operations for performing, in the same order and the same way,a finite number of steps (Baruk, 1995). More precisely in this case, it is asystematized operation for converting data into a digital format. An algorithmthus enables a large number of social security user identities to be generatedmore easily than by carrying over the forenames and surnames used forregistering civil status. Above all, it protects against homonymy, or the risk ofone individual being confused with one or several others. It also guarantees thatany newly attributed number does not correspond to any previously insuredindividual. For effective social security administration, there must be a singlenumber per individual and a single individual per number.

Finally, the choice of the algorithm may be read as a sign of certain culturaltendencies in the relationship between the State and its citizens. While certainnational mechanisms suggest great confidence in the State’s neutral and secureuse of identifiers, others betray greater caution.

Some national social security systems have seen fit to use an algorithm thatdraws on the personal lives of individuals. This is particularly the case for theFrench social security number. Composed of thirteen digits, it enables immediateidentification of the gender of the individual; their month, year and place ofbirth; and their “birth order”, which allows for the “identification of personsborn at the same place in the same period”.8

At the other end of the spectrum, the social security identifier can assume a lessmeaningful form or even resemble a random string of characters. This notablyapplies to the United States’ Social Security number. In fact, its first three

8. See “Les règles d’identification et d’immatriculation des assurés par les organismes sociaux”,<www.ameli.fr/assures/droits-et-demarches/index.php>.




7

http://www.ameli.fr/assures/droits-et-demarches/index.php

characters – the “area number” – originally corresponded to a geographicalordering of the local social security offices responsible for registering eachworker. The “group number” and the “serial number” correspond more to anadministrative registration order, although the classification used for the “groupnumber” follows relatively complex sequencing.9 Moreover, since 2011, theUnited States Social Security Administration has been following a “randomization”programme that aims to remove any possibility of reading a geographical referenceinto any assigned number. Obviously, this programme will never make thegeneration of numbers completely random. Nevertheless, the algorithm it appliesenables the twin principles of one number per insured person and one insuredperson per number to be retained.

A secure device

Algorithmic rules do not merely provide an order for generating and assigningregistrations to insured persons. Most importantly, they are also powerfulinstruments for preventing errors and fraud.

In addition to ensuring that there are no redundant identifiers, these rulesenable the coherence of each insured person’s number to be verified. The factis that each and every communication of the number carries a risk of are-transcription error by the administration. Whether the communication is oralor written, on a form, at a counter, by post or by email, this risk is alwayspresent. Even though errors occur in only a small proportion of operations, thismay equate to a considerable number of errors in a system covering severalmillion insured persons. This risk must not, therefore, be neglected. Bydefinition, the algorithm produces a logical regularity that makes the form ofeach number predictable. It follows that any sequence of characters can be testedby the algorithm in order to verify that it could indeed have been generated by it.Of course, the exact form this test takes will vary according to the algorithm used.

In France, the verification procedure may relate to the coherence between thelast group of two figures and the preceding figures. For this reason, the last twofigures are called the “key”. In order to be valid, this key must, when added tothe first 13 figures, equal a multiple of 97 (Lévy, 1989). In Canada, the verificationrelies on a substantially different algorithm, which nevertheless fulfils a similarfunction. It is presented as an application of the Luhn formula.10

9. See “Social Security Numbers. The SSN Numbering Scheme”, <www.ssa.gov/history/ssn/geocard.html>.10. This algorithm bears the name of its inventor, who filed the patent in the 1950s. On this subject, seeLuhn (1960); see also Auditor General of Canana (2002) and Houle (1986).




8

https://www.ssa.gov/history/ssn/geocard.html

https://www.ssa.gov/history/ssn/geocard.html

Although the term “key” refers to an intangible key, it is in fact nothing morethan a device intended to verify that its holder is authorized to open,metaphorically, the door fitted with the corresponding lock (Latour, 1996). Theverification key of the social security number is no more infallible than a doorkey: an existing number can always be fraudulently used by an individual who isnot insured. But, at the very least, it is a mechanism that is relatively easy to putin place to protect against many errors and the most basic forms of fraud.

In addition to the algorithm used for social security number generation,national social security systems possess important mechanisms to make the useof the identifier secure. As discussed, in some national legislations the numbersometimes directly conveys certain personal details. In other systems, the initialdata are immediately made anonymous or encrypted in the composition of thenumber. This last technique should be recognized as an important safeguardagainst the most basic attempts to invade the privacy of individuals. For thisreason, it seems more modern and relevant. But no matter which technique isused, a social security number is by definition essential to the functioning ofinformation systems that link numbers to personal data. The data held willcorrespond, as a minimum, to the conditions required to permit an individual tobenefit from social security rights.

In this light, the social security number is a sensitive piece of data from the pointof view of individual rights and liberties. It is a technical mechanism that mayenable a public entity or third party to collect the personal data of insuranceusers without their knowledge and to their cost, which means that it may beused for significant fraud or wrongdoing. The personal nature of the data heldtherefore necessitates special protective measures. This need is of course reflectedby legal rules, but also by technical mechanisms. With regard to legal rules,international and national laws aim to strictly monitor the computerizedcollection of personal data.11 This monitoring is not limited to the social securitysystem, albeit that this is considered a fundamental area. With regard to thetechnical mechanisms, protective hardware and software are crucial to ensuringthe practical implementation of data security. The most widespread technique isto make personal information anonymous. Once again, it is the use ofencryption algorithms that ensures the effectiveness of anonymization.

Through these technical mechanisms, the social security number enables acompromise between the need to protect the fundamental rights of theindividual and the need to monitor the operations of social security institutions.The scale of this compromise highlights the crucial normative function that thesocial security number plays in social protection systems.

11. For a synopsis of legal norms, see Stoddart (2012).




9

Significant normative power

The most patently normative and legal aspect of the social security number isregistration. Here, registration refers to social security numbers seen both as asequence of characters and as an individual’s enrolment in the ranks of thosecovered in the eyes of the law. Taken in the latter sense, the term allows us togain a clearer idea of the social security number’s normative dimension. It doesnot merely express someone’s identity in an abbreviated form; rather, itoperationalizes a means of ascertaining who is or is not an insured person. Fromthis point of view, the social security number becomes a powerful tool forensuring that the system functions properly and its benefits are awarded asintended. Its power goes even further, however, in that it represents the first steptowards extending coverage in pursuit of the ultimate goal: universal protection.

The gatekeeper

From the point of view of registration, the social security number is nothing lessthan the device that enables a social security system to function. It acts as a kindof automated entry and exit system, a gatekeeper, at the legal perimeter of thesocial security system.

An individual’s registration is associated with a panoply of social rights viewedas an indivisible whole. A social security number therefore grants to insuredpersons access to the social security benefits to which they are entitled. It is adevice for demonstrating a person’s identity and legal status to the relevantinstitutions. Although a social security number is not an identity document initself – it is in fact legally insufficient as proof of an individual’s identity12 – it isa direct manifestation of that identity and therefore does have some legal validity.Institutions can use it to determine whether financial support, a pension, medicaltreatment or other social security benefit should be awarded or denied, and itallows the granting of the correct benefit to the correct person.

From the point of view of the insured, without a social security number, benefitswould not be accessible even if, theoretically, they were eligible to receive them.While the social security number may be viewed as a minor tool in a socialsecurity system, in practice its role is pivotal. Due to the fact that it enables aninsured person to register as one of a group covered by the system, it is thisnumber which allows the individual to “exist” in the eyes of social security law.

12. To be more precise, only a social security “card” could, in some circumstances, be deemed to be anidentity document. In the legal systems addressed in this article, social security cards are not afully-fledged form of identification if they do not feature a photo of the insured person.




10

This is the key function of the social security number: to serve as a meansof according rights, which is a prerequisite for accessing those rights. This is, ofcourse, only by virtue of the fact that the number corresponds to personalinformation that makes an insured person eligible for social security benefits.Therefore, it does not perform its role in isolation, but rather only by beingconnected to the legal criteria for eligibility in a given social security programme.Generally, it is an important, if not crucial, device for establishing andmaintaining any social security system.

Besides serving to ensure effective access to social security protection, thenumber is also a powerful means to ensure that benefits are awarded inaccordance with the law. From this perspective, a social security number nolonger seems merely to be an approximate expression of an insured person’sidentity; it appears, rather, to constitute an actual component of a social identity.In certain cases of fraud, that identity can be usurped. So-called identity theft issometimes detrimental to the insured person involved, such as in cases wherebenefits to which they are entitled are dispensed to a third party. It can also bedetrimental to the institution that has been led to grant a benefit erroneously.

Social security registration systems also make a number of other checks possible.First, because the benefits awarded to a given number can be checked against theinsured person’s legal status, the social security number can be used to ensurethat a benefit has been granted to the correct person. Second, registrationsystems make it possible to ensure that each number is used by only one person.Computer-based systems even make it possible, by means of cross-checking, totest remotely whether the benefits granted to a particular person are credible andeven to check that one person is not using several numbers in order to claimbenefits to which they are not entitled.

Finally, in addition to the fraud scenarios mentioned above, not to beoverlooked is the fact that the risk of material error is always considerable whenworking with systems covering millions of insured persons. In other words, thesocial security number as a technique and the extent to which it is reliable willdetermine how efficiently error and fraud can be combatted and, in broaderterms, how reliable the entire social security system is.

The financial controller

The technique of registration also makes it possible to define the extent of thegroup of insured persons. Such a function is not strictly inherent to the conceptof social security; rather, it indicates the manner in which the institution shouldbe managed. In this sense, the social security number is the system’s financialcontroller: it holds the purse strings.




11

Bymaking it possible to ensure that rights are applied in accordance with the law,a social security number is also the main device allowing those individuals who areprotected by the relevant institutions to be distinguished from those who are not, orat least not yet. In other words, this device ensures that a given social security systemis not universal, even though that may be the ultimate aim.

On the surface, it would seem entirely possible to conceive of a social securitysystem in which social security numbers did not exist. We might even go so faras to imagine a legal system in which many rights could be exercised without anyneed for a number or proof of identity. In this case, the system would simply beuniversal and unconditional. Benefits would be granted to anyone requestingthem on the sole basis of their declared need.

While this thought process is perfectly coherent, implementing it wouldnaturally lead to a reality in which financial resources were limited. Nor is thisreality in any way derived from the funding modalities used in a given socialsecurity system. Whether they are sourced from the State, contributors, taxes orother types of social contributions, sources of finance are always limited.

Ultimately, then, the social security number is not a mere numerical key to eachinsured person’s identity; it is a primary material means of controlling socialsecurity expenditure. It is what prevents the system from being open to all,which makes it not only its lock and key and its gatekeeper, but also its financialcontroller. This management function is not only primordial, it is permanent. Itmakes it possible not only to ascertain who is entitled to what in the socialsecurity system, but which institution is to cover each item of expenditure in fullor in part. As an accounting tool, then, the social security number is a powerfulmeans of assessing and analysing how the social security system is functioning.

In turn, besides enabling a quantitative assessment of institutional activity, socialsecurity numbers are the first step towards a more qualitative appraisal in that theallocation of social security benefits generates information that can be used instatistical studies to shed new light on the precise circumstances of beneficiarycohorts. Such statistics may of course reveal epidemiological trends, but alsoinsured persons’ economic and social circumstances.

A primary tool for coordinating coverage

Social security numbers facilitate coordination among institutions to ensurecontinuity of coverage. From the outset, the world’s first social security systemswere presented as means of bringing greater security to workers and addressingeconomic and social risks across the life course. They also offered an efficientway of providing employers with a healthier and more stable workforce. Now, inthe most affluent countries, this protection tends to benefit all sectors of the




12

population. As such, registration ensures that insured persons are protected in acoherent and uninterrupted manner regardless of any changes to theircircumstances. Thus, an individual may, for example, receive a number at birththat will “accompany” them when they become an adult, find a job, lose theirjob and draw their retirement pension, and so on. The insured person willtherefore keep the same number from initial registration to death – and in somecases that number may continue after a person’s death to be used to determinebenefit entitlements. This continuity means that, within a single social securitysystem, an individual can access rights reserved for distinct sectors of thepopulation, such as students, employees, self-employed workers, pensioners, etc.Social security numbers are therefore a technique allowing a single system to dealwith changes to an individual’s legal status over time.

For access to a social security programme that operates within a fragmentedsocial security institutional structure, the fact that each insured person has asingle social security number enables coordination between multiple institutions.In the field of health care, for instance, an insured person may be treatedby several health care professionals or establishments, which may be funded byseveral different institutions (public facilities, social security institutions,cooperatives, private insurers, etc.). On each occasion, the insured’s health carecosts can be linked to the same number.

Moreover, having a single social security number can do much to facilitatecontinuity of coverage for an individual travelling abroad. For many decades,the institutions of the European Union (EU) have worked on the basis that theefficient transmission of social security information is beneficial for workermobility.13 In practical terms, the efficient transmission of social securityinformation between national jurisdictions is especially important whencalculating retirement pensions.14

A single, permanent social security number is not the only means of ensuringcontinuity of coverage. Temporary registration systems do exist, as is the case,for instance, in Canada, where temporary residents, be they workers or students,receive a number with an expiry date and which begins with the number 9.15 Inthe EU, a European health insurance card can be provided to any insured person

13. See, for example, the preamble to Council of Europe Recommendation No. R (86) 1, dated23 January 1986, on the protection of personal data for social security purposes <www.coe.int/t/dghl/standardsetting/dataprotection/dataprotcompil_en.pdf>.14. See Council Regulation (EEC) No. 574/72 of 21 March 1972 fixing the procedure for implementingRegulation (EEC) No. 1408/71 on the application of social security schemes to employed persons andtheir families moving within the Community, Article 50; and Regulation (EC) No. 987/2009 of theEuropean Parliament and of the Council of 16 September 2009 laying down the procedure forimplementing Regulation (EC) No. 883/2004 on the coordination of social security systems, Article 52.15. Social Insurance Number Regulations, 23 April 2013, SOR/2013-82, Sections 8 and 9 <laws-lois.justice.gc.ca/eng/regulations/SOR-2013-82/FullText.html>.




13

https://www.coe.int/t/dghl/standardsetting/dataprotection/dataprotcompil_en.pdf

https://www.coe.int/t/dghl/standardsetting/dataprotection/dataprotcompil_en.pdf

http://laws-lois.justice.gc.ca/eng/regulations/SOR-2013-82/FullText.html

http://laws-lois.justice.gc.ca/eng/regulations/SOR-2013-82/FullText.html

residing in an EU country other than that in which they enjoy health insurancebenefits.16 This card bears a specific number that is recognized by all nationalsystems within the EU, but which is granted temporarily and is valid only up tothe expiry date on the card.

Elsewhere, there are cases where more than one numbering system exists withina single national legal system, each one corresponding to different social securitybenefits or domains. One such case is France, where registration for family benefitsand health benefits are separate. However, in all cases, the presence of multiple ortemporary registration inevitably makes inter-institutional information andcommunication systems more complex.

Such complexity may bring certain benefits, such as making it more difficult tousurp identities; the greater the extent to which information about a single insuredperson is spread across several registries, the more difficult it will be for someonewith ill intent to effectively join up all the pieces. However, such complexity alsocarries an administrative burden that makes institutions less transparent andmore difficult to access for both enterprises and insured persons. Taking steps tosimplify these procedures, such as the use of a single number, thereforepromotes access to social protection. Meanwhile, the fight against identity theftcalls for data protection mechanisms that are both adequate and painless forusers, such as data encryption of server protection software in social securityinstitutions.

Towards a universal social security system?

Universal access to social security protection is the global ambition underlyingmodern notions of social security (see, for example: ISSR, 2007; ILO, 2009 and2012; De Wispelaere and Stirton, 2012; Hagemejer and McKinnon, 2013). Asystem for the transnational identification of insured persons could be a steptowards achieving this.

Although the use of an identifier is essential for establishing and maintaining anysocial security system, the main international norms on social protection do notmake reference to it. The ILO Social Security (Minimum Standards) Convention,1952 (No. 102) mentions neither social security numbers nor registration. Nor dothe ILO Recommendation concerning National Floors of Social Protection, 2012(No. 202), ILO Convention concerning Equality of Treatment of Nationals andNon-Nationals in Social Security, 1962 (No. 118), or ILO Convention concerning

16. Administrative Commission of the European Communities, 2003/751/EC, Decision No. 189 of18 June 2003 aimed at introducing a European health insurance card to replace the forms necessary forthe application of Council Regulations (EEC) No. 1408/71 and (EEC) No. 574/72 as regards access tohealth care during a temporary stay in a Member State other than the competent State or the State ofresidence (Text with relevance for the EEA and for the EU/Switzerland Agreement).




14

the Establishment of an International System for the Maintenance of Rights inSocial Security, 1982 (No. 157). Although, in the more favourable and restrictedcontext of the EU, the technical requirement for social security numbers to becommunicated between national administrations has been explicitlyacknowledged, the only practical measure that this has borne has been for singleindividuals to be endowed with several different identifiers.

Admittedly, social security registration is a method of implementation ratherthan a right, at least in the strict sense of the term. Perhaps, then, there is nourgent need for the international community to reach a consensus on the matter.However, once the idea of a universal social security system has been mooted, itbecomes difficult not to contemplate the essential practical steps that would berequired to achieve it. How would it be possible to determine whether everyonewould benefit from effective social protection without first being able to identifyprecisely “everyone” (in this regard, see also De Wispelaere and Stirton, 2012).Would a single identifier for each insured person, recognized by all nationalsystems, not be an efficient and effective step towards universalism? As well asfacilitating the national and international mobility of the insured, a singleidentifier system could present a technical solution for registration in countrieswhere one is not yet in place and whose social security systems have yet to bebuilt or perfected. This idea warrants examination from every angle. Althoughmany countries do not have sufficient administrative capacity for such aregistration system to be truly effective, efficient and secure, these shortcomingsrelate more to the implementation of the tool than to the tool itself. This is aquestion deserving of study.

As powerful as such a tool might prove, there is a need to take into account thepotential risks to individual freedoms. Prior to addressing the technical matter ofpersonal data protection, consideration should be given to the ethical dimensionof creating a database containing identifiers for the entire world population. Therisk of such a database being misused to criminal or political ends would beproportional to the number of individuals included in it.17

However, these risks already exist to a large extent in relation to many otherdatabases. But their highly technical nature and the protection to which they aresubject appear to prevent these attracting much public attention. Meanwhile,other plans for universal identification systems are underway, and in some caseshave even been implemented. For the time being they have taken the route ofinternational technical normalization. This is true of bank identifiers in the formof the IBAN and BIC norms. Other norms on technical identifiers exist and arein use for bibliographical sources and digital objects. As for human identifiers,

17. On the misuse of personal data to political ends, see, for example, Seltzer and Anderson (2001); andNader (2001).




15

ISO/IEC standard 24703:2004 defines the information that may be associated withparticipants in e-learning programmes.18

Conclusion

A move to create a universal social security identifier may present an opportunityto reawaken public interest in the way in which identifiers are formed and used.Technical normalization may be a credible path towards achieving this, but thisapproach currently lacks transparency and is thus not yet adapted to the politicaland social context. In contrast, discussion of the technical methods used toidentify and monitor insured persons should include civil society to the greatestextent possible. An exhaustive comparative study of the benefits and drawbacksof each technical option should be carried out as a basis for debate, and evenmore importantly, if the project proceeded, it would have to do so with thehighest level of transparency and public scrutiny.

The project should also be assessed on the basis of whether such a measurewould truly give rise to new rights. This is not certain to be the case ifconsumers are increasingly unaware of the expertise that goes into designing andproducing the technical processes that surround them in their daily lives. Theday when the universal right to social security is truly effective will come wheninsured persons are able to understand the technical processes used to theirbenefit. Only then will they understand the meaning and import of the numbersthat appear on their administrative documents; only then will they be trulyequipped to understand precisely how their social security system works.

Bibliography

Auditor General of Canada. 2002. “Chapter 1: Human Resource Development Canada:

The integrity of the social insurance number”, in Report of the Auditor General of

Canada. Ottawa, Office of the Auditor General of Canada.

Baruk, S. 1995. Dictionnaire de mathématiques élémentaires: pédagogie, langue, méthode,

exemples, étymologie, histoire, curiosités. Paris, Seuil.

Béchillon, D. 1997. Qu’est-ce qu’une règle de droit. Paris, O. Jacob.

Council of Europe. 1986. Recommendation No. R (86) 1 of the Committee of Ministers to

Member States on the protection of personal data used for social security purposes (Adopted

18. The full title is: ISO/IEC 24703:2004 Information technology – Participant identifiers <www.iso.org/iso/catalogue_detail.htm?csnumber=38921>. On this norm and the surrounding debate, see, forexample, Perriault, Arnaud and Juanals (2002).




16

http://www.oag-bvg.gc.ca/internet/docs/20020901ce.pdf

http://www.oag-bvg.gc.ca/internet/docs/20020901ce.pdf

https://rm.coe.int/CoERMPublicCommonSearchServices/DisplayDCTMContent?documentId=09000016804dd352


http://www.iso.org/iso/catalogue_detail.htm?csnumber=38921

http://www.iso.org/iso/catalogue_detail.htm?csnumber=38921

by the Committee of Ministers on 23 January 1986 at the 392nd meeting of the Ministers’

Deputies). Strasbourg, Council of Europe – Committee of Ministers.

De Wispelaere, J.; Stirton, L. 2012. “A disarmingly simple idea? Practical bottlenecks in the

implementation of a universal basic income”, in International Social Security Review,

Vol. 65, No. 2.

Goyard-Fabre, S. 1986. “Le droit est-il de ce monde?”, in Droits, No. 4.

Griffe, M. 2006. De la carte perforée à l’ordinateur et Internet (Les chronologies de Maurice

Griffe, No. 49). Le Cannet, Tableaux synoptiques de l’histoire.

Hagemejer, K.; McKinnon, R. 2013. “Introduction: The role of national social

protection floors in extending social security to all”, in International Social Security

Review, Vol. 66, No. 3–4.

Houle, Y. 1986. Les contrôles comptables et la vérification en milieu informatique. Québec,

Presses de l’Université Laval.

Hurtaut, P. 1999. De la carte perforée à l’ordinateur: vingt années chez “Bull”. Lauris, Editions

du Lubéron.

ILO. 2009. Social health protection: An ILO strategy towards universal access to health care

(Social security policy briefings, No. 1). Geneva, International Labour Office – Social

Security Department.

ILO. 2012. Social security for all: Building social protection floors and comprehensive social

security systems – The strategy of the International Labour Organization. Geneva,

International Labour Office – Social Security Department.

ISSR. 2007. “Extending social security to all”, in International Social Security Review,

Vol. 60, No. 2–3 (entire issue).

Kelsen, H. 1962. Théorie pure du droit. Paris, Dalloz.

Latour, B. 1996. “La clef de Berlin et autres leçons d’un amateur de sciences”, in Petites

leçons de sociologie des sciences. Paris, La Découverte.

Lévy, M. L. 1989. “Le numéro d’identité”, in Population et sociétés, No. 232.

Luhn, H. P. 1960. Computer for verifying numbers US2950048 A (patent).

Nader, R. 2001. “Opt-out for your privacy”, in The Nader Page, 21 June.

Perriault, J.; Arnaud, M.; Juanals, B. 2002. “Les identifiants numériques humains:

éléments pour un débat public”, in Les Cahiers du numérique, Vol. 3, No. 2.

SEMA. 1969. La carte perforée: cours programmé par le Service d’enseignement programmé de

la Sema. Paris, Dunod.

Seltzer, W.; Anderson, M. 2001. “The dark side of numbers: The role of population data

systems in human rights abuses”, in Social Research, Vol. 68, No. 62.

Stoddart, J. 2012. International privacy standards: Development, recent events and limitations

(Remarks at the 43rd Annual Study Session of the International Institute of Human

Rights, 9 July). Ottawa, Office of the Privacy Commissioner of Canada.




17



http://onlinelibrary.wiley.com/doi/10.1111/j.1468-246X.2012.01430.x/abstract

http://onlinelibrary.wiley.com/doi/10.1111/j.1468-246X.2012.01430.x/abstract

http://onlinelibrary.wiley.com/doi/10.1111/issr.12016/abstract

http://onlinelibrary.wiley.com/doi/10.1111/issr.12016/abstract

http://www.ilo.org/wcmsp5/groups/public/---ed_protect/---soc_sec/documents/publication/wcms_secsoc_5956.pdf

http://www.ilo.org/secsoc/information-resources/publications-and-tools/books-and-reports/WCMS_SECSOC_34188/lang--en/index.htm

http://www.ilo.org/secsoc/information-resources/publications-and-tools/books-and-reports/WCMS_SECSOC_34188/lang--en/index.htm

http://onlinelibrary.wiley.com/doi/10.1111/issr.2007.60.issue-2-3/issuetoc

https://www.scribd.com/doc/61470451/Theorie-pure-du-droit-1962-Hans-Kelsen

http://www.ined.fr/fr/publications/population-et-societes/le-numero-d-identite

https://www.google.com/patents/US2950048

https://blog.nader.org/2001/06/21/opt-out-for-your-privacy/

https://www.cairn.info/revue-les-cahiers-du-numerique-2002-2-page-169.htm

https://www.cairn.info/revue-les-cahiers-du-numerique-2002-2-page-169.htm

https://pantherfile.uwm.edu/margo/public/publications/darkside4897395.pdf

https://pantherfile.uwm.edu/margo/public/publications/darkside4897395.pdf

https://www.priv.gc.ca/en/opc-news/speeches/2012/sp-d_20120709/

Survivors’ pensions and theirimpact on the Brazilian

labour marketRogerio Nagamine Costanzi*, Graziela Ansiliero* and

Julimar Da Silva Bichara**

*IPEA, Brasilia, Brazil; **Universidad Autónoma de Madrid, Spain

Abstract In many countries the regulations governingsurvivors’ pensions were established in periods when femalelabour market participation was lower than at present.However, the current trend in many Latin Americancountries is for growing levels of female labour participation.In Brazil, where there are no restrictions on the concurrentreceipt of retirement and survivors’ pensions, and whereuntil recently lifelong pensions could be obtained withoutany conditionality, not only has the accumulation of suchbenefits grown, but there are indications that these rules havehad a negative impact on women’s labour marketparticipation. Analysis of the case of Brazil shows the needfor social security regulations to adapt to labour marketchanges, and underlines the need to acknowledge that socialsecurity regulations can actually have an impact on the labourmarket.

Keywords survivors benefits, social security reform, labourmarket, woman worker, Brazil

Addresses for correspondence: Rogerio Nagamine Costanzi, Instituto de Pesquisa Econômica Aplicada(IPEA), SBS (Setor Bancário Sul) Edificio BNDES sala 1405, 70076–900, Brasília – DF, Asa Sul, Brazil;email: [email protected]. Graziela Ansiliero, Instituto de Pesquisa Econômica Aplicada(IPEA), Diretoria de Estudos e Políticas Sociais (Disoc), SBS (Setor Bancário Sul) Edificio BNDES,sala 1408, 70076–900, Brasília – DF, Asa Sul, Brazil; email: [email protected]. Julimar daSilva Bichara, Depto. Estructura Económica y Economía del Desarrollo, Universidad Autónoma deMadrid, Ciudad Universitaria de Cantoblanco, C/ Francisco Tomás y Valiente, 5, 28049 Madrid,Spain; email: [email protected].



19

Introduction

Within Brazil’s General Social Security Scheme (GSSS), the survivors’ pension(widows and orphans)1 consists of a cash benefit granted to the spouse and/ordependants of deceased persons where they meet the requirements established bythe social security system. Until very recently, the rules for the award andretention of survivors’ pensions under the GSSS were commonly regarded asbeing very flexible, a situation that led to the drafting and implementation ofseveral parametric adjustments to the regulations governing this type of benefit.

Across 2014–15, a reform of the regulations for survivors’ pensions wasapproved by Provisional Measure (PM) No. 664 of 30 December 2014, whichbecame Act No. 13,135 of 17 June, 2015. This reduced the differences betweenBrazilian and international rules with respect to this important type of socialsecurity benefit.

Before the adoption of this reform, comparison of Brazil’s system of survivors’pensions and those of 132 other countries, made in national reports compiled byboth the International Social Security Association (ISSA) and the InternationalLabour Organization (ILO), showed the almost total absence of conditionsgoverning access to this type of benefit in Brazil. The only requirement was tohave been a member of the social security scheme. Presently, it is still possibleunder certain circumstances to obtain benefits with the insured only having paida single contribution (Ansiliero, Costanzi and Pereira, 2014).

One of the reasons for the reform – but not the only one – was the growingoccurrence of the accumulation of retirement pensions and survivors’ pensionsby widows and widowers. Other factors that influenced the decision to changethe rules for survivors’ pensions have been the greater tax burden resulting fromthe ageing of the population and the existence of rules that permittedopportunistic behaviour. From a more general perspective, Brazil is also followingthe pattern seen in other countries that have introduced counter-cyclical measuressince the crisis of 2008–09, with measures subsequently being taken to achievefiscal adjustment and to improve the country’s public finances, which haddeteriorated significantly in 2014 and 2015.

However, from the point of view of social policy, the discussion is centred onthe search for sustainable social security policies. As will be shown, theaccumulation of benefits by individuals has a regressive effect on incomedistribution in Brazil. Specifically, the accumulation of retirement pensions andsurvivors’ pensions without restrictions mainly benefits individuals with thehighest incomes.

1. Hereafter referred to as survivors’ pensions.

Survivors’ pensions and the Brazilian labour market



20

To evaluate the changes in the rules governing survivors’ pensions and the logicunderlying these in Brazil, this article is organized as follows: the next sectionprovides an analysis of developments in the accumulation of retirement benefitsand survivors’ pensions, and an evaluation of its effects on the Brazilian labourmarket. The article then describes the Brazilian reform and offers a brief analysisof the changes in the rules for survivors’ pensions. The article concludes by offeringa number of final considerations.

Number of survivors’ pension beneficiariesand the pattern of coverage in Brazil

It is important to start this discussion by emphasizing that the survivors’pensions rules in Brazil were originally formulated at a time when women’sparticipation in the labour market was lower than at present and women’sprotection was often linked to economic dependency relationships (usuallywith parents or husbands covering their needs). However, over the past fewdecades the country has experienced significant changes in the labour market,notably in women’s economic activity and the pattern of their entry into thelabour market.

With regard to the pattern of female labour insertion, women tend to experiencemore unemployment, even where their labour market participation is less frequent,and tend to receive lower incomes, either because of discriminatory factors orbecause they are more likely to occupy jobs that are less valued in the Brazilianeconomy and society. Thus, from the woman’s point of view, more flexiblepension rules may be regarded as (just and necessary) compensatory mechanismsfor such asymmetries and inequalities in the labour market.

Over recent decades, however, the growth in the female participation rate hasbeen considerable, and this helps to explain the evolution of unemploymentrates for men and women (Figure 1A). As shown, unemployment remainshigher among women (even with the upward trend in the rates of economicactivity for women until 2011), but with lower participation rates than men.Nonetheless, the growth in female economic activity (Figure 1B) is significantand helps to explain the evolution of the differences observed between menand women.

In terms of total income, for all jobs declared in the National HouseholdSurvey (IBGE, various years) and based on the hours worked by men andwomen regardless of the nature of the occupation and the sector of economicactivity in which the job is performed, it may be observed that the difference inremuneration between men and women has declined over the last two decades(Figure 2). The differences in average remuneration between the sexes, althoughstill present, also fell significantly over the same period of time.




21

Figure 1A. Brazil: Unemployment rates, ages 16–59, by sex – harmonized series(1992–2014) (percentages)

Figure 1B. Brazil: Rates of economic activity of employed persons, ages 16–59, by sex –harmonized series (1992–2014) (percentages)

Note: In the series with data from 1992 to 2003, the rural areas of the North region, except for Tocantins, are notincluded.

Source: IBGE (various years). Authors’ interpretation.




22

Brazilian women have thus achieved greater labour market participation andhave done so in better conditions, taking advantage of the economic upturnexperienced by the country (at least until 2014, the year of the most recentNational Household Survey available; see IBGE, 2014). The percentage of theemployed population between ages 16 and 59 who contribute to social securityhas grown significantly over the last decade. This figure, which already signalled aconvergence between the sexes, was in 2013 and 2014 slightly higher amongwomen (statistically, indicating the gender balance) (Figure 3). This means that,although gender equality has not yet been achieved, at least the pattern of femalelabour market participation has changed significantly, and Brazilian socialsecurity should reflect this change.

With changes in thepatternofwomen’s labourmarket participation, it is necessaryto rethink the rules applying to survivors’ pensions. In 1992, 9.9 per cent ofBrazil’s pensioners2 accumulated pension benefits both for survivors’ pensionsand their own pensions (old age or disability), which rose to 31.6 per cent in 2013(see Table 1). The trend is for this percentage to rise higher still due to the

2. Excluding the rural northern area of the country (with the exception of the State of Tocantins).

Figure 2. Brazil: Evolution of wage disparities by sex – Ratio of men’s to women’shourly remuneration, all professions, harmonized series (1992–2014) (percentages)

Note: In the series with data from 1992 to 2003, the rural areas of the North region, except for Tocantins, are notincluded.





23

increasing rate of female participation. Taking into account women only, whoconstitute the majority of pensioners, the proportion of those accumulatingsurvivors’ pensions with retirement pensions increased from 9.9 per cent in 1992 to31 per cent in 2013.

Women in 2013 represented about 87 per cent of survivors’ pensionbeneficiaries in the country; in other words, the vast majority (albeit that thisfigure is lower than that observed in 1992, when it was 93.4 per cent). Oneexplanation for this high figure is that women have a longer life expectancy thanmen, and that the average age of women at marriage is lower than that of men.In general, men marry younger women. The average age of men at marriage rosefrom 27 to 30 years between 1974 and 2014, while among women it rose from23 to 27 years during the same period.3

3. This rise in the average age at marriage in recent years may reflect the greater determination ofindividuals to seek to earn higher wages, although it may also stem from the increasingly commondeferral of formal marriage. Regardless, the observation remains valid that men tend to marryyounger women.

Figure 3. Brazil: Evolution of pension coverage for persons aged 16–59, by sex –Proportion of employed persons contributing to social security (RGPS and/or RPPS)(2004–2014) (percentage)

Note: RGPS: Social General Social Security Scheme (Regime Geral de Previdência); RPPS: Special SocialSecurity Scheme (Regime Próprio de Previdência Social).





24

Table 1. Situation of survivors’ pension beneficiaries (accumulation of benefits, statusof activity, employment status, and sources of income) by gender, 1992 and 2013 –Brazil

Brazil – Men and Women

Pensioner’s situation 1992 excludingthe northernrural region

2013 excludingthe northernrural region

2013 includingthe northernrural region

I – Total number of pensioners (II + III) 3,339,086 6,903,831 6,954,563

II – Pensioners accumulating survivors’and retirement pensions

330,046 2,179,566 2,196,235

II.1 Unemployed 1,725 1,618 1,618

II.2 Employed, contributing to social security 12,943 37,215 37,459

II.3 Employed, not contributing to social security 41,885 304,488 309,937

II.4 Economically inactive 273,493 1,836,245 1,847,221

III – Survivors’ pension beneficiaries notalso receiving retirement pensions

3,009,040 4,724,265 4,758,328

III.1 Unemployed 66,365 79,119 79,362

III.2 Employed, contributing to social security 382,386 562,952 565,357

III.3 Employed, not contributing to social security 638,821 628,667 639,690

III.4 Economically inactive 1,921,251 3,453,527 3,473,919

III.5 Situation not stated 217 - -

% accumulation survivors’ and retirement pensions(II/I as % of total)

9.9 31.6 31.6

% who have another formal source of income 21.3 39.7 39.7

% who have another source of income 40.5 48.8 48.9

Brazil – Women




I – Total number of pensioners (II + III) 3,118,255 6,007,517 6,044,737

II – Pensioners accumulating survivors’and retirement pensions

307,690 1,863,030 1,875,895

II.1 Unemployed 1,470 1,618 1,618II.2 Employed, contributing to social security 10,819 28,100 28,100

II.3 Employed, not contributing to social security 37,366 244,471 248,973

II.4 Economically inactive 258,035 1,588,841 1,597,204(Continued)




25

It is important to note that this percentage is not higher because there arewomen who receive survivors’ pensions while they have not yet reachedretirement age.4 Taking into account survivor pensioners aged 60 or older, whonumbered around 4.5 million in 2013, about 2 million accumulated pensions,which represented approximately 45.3 per cent of the total. This percentage hastended to grow due to the increase in female participation in the labour market.In 1992, the respective percentage was 16.6 per cent.

Therefore, it is also important to take into account survivors’ pensionbeneficiaries who also receive retirement pensions, as well as those pensionerswho were employed and contributing to social security. In this case, thepercentage of survivor pensioners with at least two formal sources of incomeincreased from 21.3 per cent in 1992 to 39.7 per cent in 2013 (see Table 1;sum of II + III.2 of the total number of pensioners). With regard to women,that percentage increased from 20.6 per cent in 1992 to 38.6 per cent in 2013.Taking into account also those pension beneficiaries who had another sourceof income, including from working in the informal labour market, almost half ofthe pensioners had another source. In addition, in 2013 there were about

4. In Brazil, the normal retirement age is age 65 for men and age 60 for women.

Table 1. Situation of survivors’ pension beneficiaries (accumulation of benefits, statusof activity, employment status, and sources of income) by gender, 1992 and 2013 –Brazil - Continued

Brazil – Men and Women




III – Survivors’ pension beneficiariesnot also receiving retirement pensions

2,810,565 4,144,487 4,168,842

III.1 Unemployed 57,814 64,627 64,870

III.2 Employed, contributing to social security 334,783 454,714 455,687

III.3 Employed, not contributing to social security 586,169 515,771 522,898

III.4 Economically inactive 1,831,582 3,109,375 3,125,387

III.5 Situation not stated 217 - -

% accumulation survivors’ and retirement pensions(II/I as % of total)

9.9 31.0 31.0

% who have another formal source of income 20.6 38.6 38.6

% who have another source of income 39.4 47.2 47.2

Source: Authors’ interpretation, based on microdata from IBGE (1992; 2013a).

Brazil – Women




26

347,000 people who accumulated retirement and other pensions as well as incomefrom work, whereas in 1992 this group numbered 55,000 (see Table 1; sum ofII.2 + II.3).

Given that there is a time lag between the growth of female participation in thelabour market and the percentage of those accumulating survivors’ and retirementpensions, the trend will be for this percentage to grow further in the future. Fromthe point of view of social protection, it is clear that it seems unjustifiable to offerunconditional guarantees of income to those who already have another adequatesource of formal income. In other words, it is necessary to rethink thefundamental presumption that treats women as necessarily economicallydependent on their husbands, especially in the current scenario of women’sincreased participation in the labour market and the fact that the social securitysystem should not need to guarantee income for those who already have anotheradequate source.

It should be remembered that the survivors’ pension is a contributory benefit. Inaddition, it should be clear that financing from different sources (the personalpension contributions of the retirement beneficiary and the deceased’scontributions in the case of the survivors’ pension) is not necessarily sufficient toachieve an actuarial balance between contributions and benefits, especiallybecause of the rapid increase in life expectancy that is occurring in Latin America.

Impact on the labour market

Among the possible distortions generated by the unconditional provision ofsurvivors’ pensions, emphasis should be placed on the disincentives toparticipation in the labour market for spouses, especially widows of working agewho are fully capable of exercising paid productive activities and do not havedependant children. The lower degree of participation by women in theworkplace has always been considered solely and exclusively as a sign ofdiscrimination against women. However, in Brazil, where a survivors’ pensioncan still be paid for life to young women with full work capacity, there areindications that the lower degree of female participation in the labour market isalso determined by public social security policy.

As can be seen in Table 2, in 2013, while the rate of participation in the generallabour market of those not receiving any survivors’ pension and/or retirementpension reached 65.9 per cent, that of pensioners represented only 23.5 per cent; inthe case of women, these percentages reached 56.1 per cent and 21.9 per centrespectively, below the country’s overall average. It should be noted that in the agegroups that should be characterized by high levels of participation (ages 28–43),the proportion of economically active women exceeded 70 per cent for




27

non-beneficiaries; in the case of pensioners, this figure is just over 50 per cent, a notinconsiderable difference.

These data are a clear indication that inadequate regulations for the award andmaintenance of survivors’ pensions can result in generating disincentives towomen’s participation in the labour market, with the outcome that people whocan contribute to the country’s economic and social growth instead becomedependent on social security benefits. In this case, the difference in the

Table 2. Participation rate, proportion of employed persons and proportion ofemployed contributors among total numbers of survivors’ pension beneficiaries andtotal pension beneficiaries (excluding retirement and survivors’ pensions), by genderand age group – Brazil (2013)

Beneficiaries (survivors’ pensions)

Age group Participation rate (%) Proportion of employed persons (%) Proportion of employed contributors (%)

Men Women Total Men Women Total Men Women Total

10–15 8.6 9.4 9.0 2.6 6.3 4.5 0.0 2.2 1.1

16–21 36.0 33.0 34.2 18.9 23.0 21.3 6.4 12.7 10.2

22–27 22.3 39.5 32.9 19.9 28.0 24.9 6.4 19.4 14.5

28–32 45.5 51.4 50.1 44.1 48.0 47.1 33.3 26.7 28.3

33–38 42.4 49.8 48.2 40.5 45.2 44.2 19.3 24.1 23.1

39–43 52.3 54.1 53.8 52.3 51.2 51.4 17.3 29.8 27.5

44+ 33.8 19.4 21.1 33.2 18.7 20.4 13.1 6.5 7.2

Total 34.2 21.9 23.5 32.6 20.8 22.3 13.1 8.0 8.7

Non-beneficiaries (excluding retirement and survivors’ pension beneficiaries)

Age group Participation rate (%) Proportion of employed persons (%) Proportion of employed contributors (%)

Men Women Total Men Women Total Men Women Total

10–15 9.4 5.4 7.4 8.2 4.2 6.2 0.3 0.2 0.3

16–21 62.7 46.2 54.6 53.8 36.2 45.2 26.0 20.1 23.1

22–27 89.7 69.9 79.7 83.2 60.9 71.9 55.5 42.2 48.7

28–32 94.0 73.0 83.2 89.7 66.3 77.6 61.2 46.3 53.5

33–38 95.5 73.6 84.2 92.4 68.6 80.0 61.8 47.1 54.2

39–43 95.2 73.8 84.1 92.3 69.7 80.6 60.7 45.8 53.0

44+ 90.2 57.3 73.1 88.1 55.1 71.0 54.4 34.7 44.1

Total 76.2 56.1 65.9 72.2 51.1 61.4 44.9 33.1 38.9

Source: Author’s interpretation, based on microdata from the IBGE (2013a).




28

proportion of economically active men and women would be explained not onlyby issues related to gender discrimination (although this is a real matter ofconcern in the country), but would also be encouraged by the rules andincentives created by the country’s survivors’ pension system.

In addition, comparison of the participation rate of survivors’ pensionbeneficiaries who have children younger than age 21 and non-beneficiaries withchildren younger than age 21 shows a clear difference. While the participation rateof beneficiaries (spouses) who have children younger than age 21 was 51.5 per centin 2013, that for non-beneficiaries with children reached 80.2 per cent. Thisdifference is even higher when comparing the participation rates of beneficiariesand non-beneficiaries without children: 21 per cent for pensioners and 75.3 per centfor non-beneficiaries (see Table 3).

As regards those with different family status, that is, neither in the referencegroup nor a spouse, again it can be observed that the participation rate is muchlower than that of non-beneficiary pensioners: 10.8 per cent compared to47.4 per cent. Clearly, these data show a strong indication that the payment of alifelong pension ultimately generates a strong disincentive to labour marketparticipation, especially for widows, who constitute the majority of survivors’pension beneficiaries among insured persons. This is so even for young widowswith full work capacity with or without children.

It should be emphasized that the outcome observed in Brazil is related to therules for the award of survivors’ pensions, whereby a lifelong benefit is awardedwith virtually no conditions attached, even for spouses without children and withfull capacity for work.

A more convincing, though possibly ambiguous, way of demonstrating that theaward of a survivors’ pension without any attached conditions and based on apresumption of economic dependence, as applied in Brazil before the 2015 reform,affects the rate of activity of beneficiaries is to conduct an empirical experiment.This involves a logistic regression model, which models the relationship betweenone or more explanatory variables and a categorically dependent variable. Themodel makes it possible to estimate the probability that, given the set ofindependent variables, the variable response will take on the value of a certaincategorical response (for example, labour market participation =1; inactivity =0).

In the logistic regression model used here, the dependent variable describes theprobability of the occurrence of the event of interest (economic activity =1;inactivity =0) as a function of a set of independent variables: (i) age (continuousvariable, comprising values from 21 to 59 years); (ii) gender (dichotomous variable:woman =1, male =0); (iii) position in the family (reference person =1, anyother position =0); (iv) place of residence (metropolitan =1, non-metropolitan =0);(v) number of dependants younger than age 21 in the family; (vi) education(measured in number of years of education); (vii) social security status




29

Table

3.Survivors’

pension

beneficiaries,

proportion

ofem

ployed

personsandem

ployed

contributors

amongtotalnu

mbers

ofnon-

beneficiariesandsurvivors’pensionbeneficiaries,withandwithout

depend

ants(childrenyoun

gerthan

age21),by

gend

er,agegroup,

and

family

status

–Brazil(2013)

Survivors’pensionbeneficiaries

Non-beneficiaries

Beneficiariesin

thereferencegroupor

spouses–with

children

Non-beneficiariesin

thereferencegroupor

spouses–with

children

Agegroup

Participationrate(%)

Proportionofem

ployed

persons(%)

Proportionofem

ployed

contributors(%)

Agegroup


Proportionofem

ployed

persons(%)

Proportionofem

ployed

contributors(%)

Men

Wom

enTotalMen

Wom

enTotal

Men

Wom

enTotal

Men

Wom

enTotalMen

Wom

enTotal

Men

Wom

enTotal

10–15

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

10–15

69.9

16.2

18.4

69.9

11.6

14.0

0.0

0.7

0.7

16–21

46.7

58.1

56.0

46.7

45.4

45.6

0.0

30.0

24.5

16–21

93.2

40.9

51.7

85.3

31.3

42.5

42.6

14.3

20.2

22–27

31.3

43.9

42.7

31.3

27.8

28.2

0.0

14.9

13.5

22–27

97.1

59.8

72.7

93.6

51.6

66.2

58.2

30.0

39.8

28–32

78.5

55.7

59.0

78.5

52.1

55.9

70.2

26.9

33.1

28–32

97.3

67.9

79.8

94.0

61.4

74.6

62.0

38.3

47.9

33–38

64.9

56.2

57.4

64.9

52.6

54.3

29.6

27.5

27.8

33–38

97.8

71.4

82.8

95.2

66.4

78.8

63.3

43.6

52.1

39–43

69.3

58.4

59.7

69.3

56.9

58.4

23.2

32.3

31.2

39–43

97.6

73.3

84.4

95.3

69.4

81.2

64.2

44.8

53.7

44+

56.5

45.4

47.7

55.0

42.9

45.5

30.1

19.5

21.8

44+

95.2

66.4

82.2

93.2

63.6

79.9

59.7

40.2

50.9

Total

59.3

49.8

51.5

58.2

46.8

48.9

30.4

23.3

24.5

Total

96.7

66.9

80.2

94.1

61.7

76.1

61.2

38.8

48.8

(Continued)




30

Table3.

Survivors’pensionbeneficiaries,proportion

ofem

ployed

personsandem

ployed

contributorsam

ongtotalnu

mbersof

non-


depend

ants(childrenyoun

gerthan

age21),by

gend

er,agegroup,

and

family

status

–Brazil(2013)

-Con

tinued


Non-beneficiaries

Beneficiariesin

thereferencegroupor

spouses–withoutchildren

Non-beneficiariesin

thereferencegroupor

spouses–withoutchildren

Agegroup


Proportionofem

ployed

persons(%)

Proportionofem

ployed

contributors(%)

Agegroup


Proportionofem

ployed

persons(%)

Proportionofem

ployed

contributors(%)

Men

Wom

enTotalMen

Wom

enTotal

Men

Wom

enTotal

Men

Wom

enTotalMen

Wom

enTotal

Men

Wom

enTotal

10–15

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

10–15

45.8

24.0

25.9

41.9

16.8

18.9

9.0

1.1

1.8

16–21

44.6

30.2

32.4

44.6

22.4

25.8

29.7

22.4

23.5

16–21

89.8

58.3

70.2

83.3

47.7

61.1

45.4

27.6

34.3

22–27

64.1

46.2

50.9

64.1

40.7

46.8

26.3

29.0

28.3

22–27

95.5

80.0

87.8

91.5

72.2

81.9

64.4

54.5

59.5

28–32

58.8

53.6

54.6

46.3

53.6

52.1

17.4

33.5

30.3

28–32

97.1

84.4

91.5

93.9

79.1

87.4

70.5

63.8

67.5

33–38

64.3

43.0

47.4

64.3

39.9

44.9

43.1

20.2

24.9

33–38

96.6

82.3

90.8

94.2

78.3

87.7

67.5

58.7

63.9

39–43

57.6

44.3

47.8

57.6

35.3

41.2

19.0

19.9

19.7

39–43

95.6

74.3

85.2

92.7

70.2

81.7

61.1

46.5

54.0

44+

33.4

18.8

20.4

33.0

18.2

19.8

10.7

5.7

6.2

44+

88.8

52.4

68.0

86.8

50.7

66.2

52.5

31.1

40.2

Total

34.8

19.4

21.0

34.3

18.6

20.3

11.4

6.0

6.6

Total

91.8

60.9

75.3

89.2

57.5

72.3

57.7

38.1

47.2

(Continued)




31

Table

3.Survivors’pensionbeneficiaries,proportion

ofem

ployed

personsandem

ployed

contributorsam

ongtotalnu

mbersof

non-


depend

ants(childrenyoun

gerthan

age21),by

gend

er,agegroup,

and

family

status

–Brazil(2013)

-Con

tinued


Non-beneficiaries

Beneficiarieswith

otherfamily

status

(with

orwithoutc

hildren)

Non-beneficiarieswith

otherfamily

status

(with

orwithoutc

hildren)

Agegroup


Proportionofem

ployed

persons

Proportionofem

ployed

contributors(%)

Agegroup


Proportionofem

ployed

persons(%)

Proportionofem

ployed

contributors(%)

Men

Wom

enTotalMen

Wom

enTotal

Men

Wom

enTotal

Men

Wom

enTotalMen

Wom

enTotal

Men

Wom

enTotal

10–15

8.6

9.4

9.0

2.6

6.3

4.5

0.0

2.2

1.2

10–15

9.4

5.1

7.3

8.2

4.0

6.2

0.3

0.2

0.3

16–21

34.8

26.5

30.4

15.7

16.6

16.2

5.8

5.5

5.6

16–21

59.7

45.6

53.4

50.7

35.7

44.0

24.1

20.2

22.3

22–27

17.3

31.0

23.0

14.4

23.2

18.1

5.1

21.6

12.0

22–27

85.0

76.0

81.2

76.6

65.6

72.0

51.7

49.3

50.7

28–32

18.5

32.9

26.1

18.5

27.3

23.2

10.0

20.4

15.5

28–32

87.5

79.5

84.2

80.8

71.2

76.9

54.0

56.5

55.0

33–38

10.5

20.5

15.9

6.0

8.8

7.5

0.0

8.8

4.7

33–38

87.3

77.9

83.5

81.8

71.3

77.5

51.4

55.3

53.0

39–43

10.5

40.0

28.3

10.5

40.0

28.3

3.0

32.8

21.0

39–43

83.7

76.3

80.6

78.3

70.8

75.2

43.7

52.7

47.4

44+

6.7

8.3

8.1

6.3

8.1

7.9

2.8

3.1

3.1

44+

74.0

56.0

64.9

70.4

53.2

61.7

39.1

38.7

38.9

Total

13.2

10.3

10.8

8.7

9.2

9.1

3.2

4.2

4.0

Total

53.0

40.6

47.4

47.4

34.6

41.6

26.3

23.3

25.0

Notes:(1)Includes

thoseaged

10yearsandolder.(2)The

ratesandproportions

calculated

usethesamedenominator,i.e.,thetotalnum

berofpersons

inthesameagegroupandwith

the

samefamily

status,bygender.

Source:P

NAD

/IBG

E,2013.Interpretationby

SPPS

/MPS

.




32

(survivors’ pension beneficiary =1, non-beneficiary =0); and (viii) the age squared.All variables take as their source the 2013 National Household Survey microdata(IBGE, 2013).

Although the model’s design may need some improvement, possibly byintroducing additional explanatory variables, the refinement achieved is adequatefor the results obtained. As can be seen in Table 4, the status of the survivors’pension beneficiary, when modified by the other variables in the model, tends tosignificantly decrease the probability of participation in the labour market, andthe coefficient for this variable is significant from a statistical point of view. Theshifts in the coefficients are consistent with the expected relationship,and the impact of the status of beneficiary on the exercise of activity (the decisionon whether or not to participate in the labour market) is only inferior to theeffect of gender and family status, as would be expected.

The results indicate that the desirability of joining the workforce is substantiallylower among beneficiaries than non-beneficiaries. Given the already established

Table 4. Results of logistic regression – Impact of selected variables on the probability ofan individual’s joining the economically activity population (EAP) – 2013

Variables S.E. Coefficient(Log Odds)

Probabilityratio (Exp(B))

Δ% in probabilityof joining the EAP

Constant 0.112 -2.251 0.105 –

Position in the family (Head of household,or other position)**

0.019 0.572 1.772 77.2%

Sex** 0.019 -1.470 0.230 -77.0%

Beneficiary status (Yes/No)** 0.050 -1.110 0.330 -67.0%

Region surveyed(metropolitan/non-metropolitan)**

0.020 -0.257 0.773 -22.7%

Number of family dependants(children younger than age 18)**

0.018 -0.058 0.943 -5.7%

Major geographical regions (North, Northeast,Southwest, South and Centre-west)

0.008 0.079 1.082 8.2%

Years of study** 0.002 0.092 1.096 9.6%

Age2 ** 0.000 -0.003 0.997 -0.3%

Age (in years, from ages 21–59)** 0.006 0.203 1.225 22.5%

Pseudo-R2 Cox & Snell Nagelkerke

0.131 0.202

Note: Asterisks* indicate the relevance to the two models used in the Neyman-Pearson approach. One asteriskindicates that the ρ-value is below 5 per cent, and two asterisks indicate that it is less than 1 per cent.

Source: Authors’ interpretation, based on IBGE (2013a).




33

relationship between probability and possibilities, it can be said that the probabilityof a survivors’ pension beneficiary being employed or looking for work is lowerthan that estimated for similar individuals who differ only in that they are notsurvivors’ pension beneficiaries. These results, verified against other relevantobservable variables, strengthen the view that social security rules to a greatextent affect the labour market behaviour of beneficiaries.

In terms of social policy, it is worth noting that the accumulation of retirementpensions and survivors’ pensions generally benefits the higher income populationgroup and therefore has a negative effect on the distribution of income in Brazil.The IBGE (2014) microdata estimate showed that 70 per cent of those receivingold age and survivors’ pensions were among the richest 30 per cent of thepopulation (Figure 4).

In the international literature, it is more common to analyse the impact of socialsecurity on labour supply in terms of other types of benefits, such as conditionalcash transfer programmes (Barbosa and Corseuil, 2013; ECLAC and ILO, 2014;Foguel and Barros, 2010), and to overlook survivors’ pensions. There are,however, some important studies on the subject, albeit with different

Figure 4. RGPS and RPPS – Distribution of social security beneficiaries (retirementand survivors’ pensions beneficiaries), by deciles of per capita household income(PCHI) – 2014

Note: Deciles of distribution of per capita household income were developed on the basis of monthly income fromsources for all those living in households where all its members receive known income.

Source: Authors’ interpretation, based on microdata from the IBGE (2014).




34

methodological perspectives and approaches – possibly determined by differencesin the design of the policies analysed in the various countries that have suchsocial protection mechanisms.

Decoster et al. (2007), for example, demonstrate that the reform proposed inBelgium would remove disincentives to work among widows and widowers entitledto a survivors’ pension. This research considered that the system did not generateincentives for beneficiaries to enter the labour market by offering hours of workthat would bring an increase in their income from work above the minimumthreshold, given that this would eliminate the possibility of accumulating income,reducing their total income. This study hence strengthens the view that unsuitablesocial security rules can reduce the labour supply.

Studies looking specifically at the impact of survivors’ pensions on laboursupply are not very common, and those based on the general equilibriummodels of the Overlapping Generations Model (OLS) are even less so.Nishiyama (2015) applied these models to the Old-Age and SurvivorsInsurance (OASI) programme of the Social Security Administration of theUnited States (SSA). It concludes that the elimination of survivors’ pensionsand spousal benefit5 would increase the long-term female labour supply bybetween 4.3 per cent and 4.9 per cent; among men the respective figurewould be between 0.8 per cent and 1.1 per cent. It is worth noting that thewide difference between these results – much more significant among women – issaid to be linked to the different hypotheses raised concerning government fiscalpolicy.

In the United States, where the combination of survivors’ pensions with socialsecurity benefits and income from work is subject to a cap, it is possible that theeconomic activity of insured persons is affected, at least in part, by the rules ofthe social security system. In any case, Nishiyama (2015) reasserts the associationbetween such benefits and low labour force participation among beneficiaries,particularly women.6 Studying the United States’ labour market, Knapp (2014)examined a Family Savings Life Cycle Model, labour supply and decisions on theuse of pension benefits, supported by assumptions about health uncertainties,survival, and health expenditures, and estimates the impact of changes in survivors’pensions and spousal benefits on the decisions by each family member regardingwork and retirement.

5. In general terms, the Spousal Benefit is paid to spouses and/or former spouses of living insuredpersons who have contributed to the United States’ Social Security programme. In the event of thedeath of the insured person, a Survivor Benefit can be requested. In Brazil only the latter benefit isavailable, and the rules for its award and maintenance differ significantly from those applied in theUnited States.6. Gruber and Orszag (2003) provide statistically significant evidence concerning women in the UnitedStates.




35

In general terms, the author concludes that such benefits are relevant indetermining the labour supply of both sexes: their abolition would mean thatmen and women on average work longer. Evaluations by income levels show atendency to work less among those with lower incomes (penalties foraccumulating income from different sources represent an implicit taxation onwork, which especially affects women, who also tend to earn less), and to workmore among individuals with higher incomes from work (with a higher rate ofreturn on labour; substitution effect) thus overcoming the effect of the increasedvalue of the benefit in cases of non-accumulation (income effect, for example).The restrictions on the accumulation of income from different sources in theUnited States make it difficult to compare with Brazil, where no such restrictionsare currently applied. That is, the impact of the supply of these two benefits is tosome extent confused with the impact of the specific rules associated with them,which limits direct comparison with the case of Brazil.

Similarly, James (2009), in a study that analyses the rules, rationale and impactsof survivors’ pensions (in 39 countries), recognizes the potential impact ofsurvivors’ pensions on decisions by potential beneficiaries concerning work, bothwhere they already receive such transfers and where they only expect to receivesuch benefits in the future (especially where the rules of the system penalizethe accumulation of such benefits with other social security benefits, financedby the beneficiary’s own contributions). That is, certain social security rules,although formulated to reduce costs, may lead to undesirable distortions in thebehaviour of beneficiaries. Thus, when they are applied, it can be veryproblematic to identify a cause and effect relationship between conditionalityregarding activity and actual work among beneficiaries.

In several countries (including the United States), the survivors’ pension isreduced, cancelled or affected by compulsory choices (between their value andother monetary income) if the beneficiary (usually young) is receiving a wage orreceiving social security benefits (usually elderly persons). If the deceased insured(who generated the right for the payment of the survivors’ pension) receivedhigh remuneration, the dependants will tend to receive a survivors’ pension witha value higher than other benefits based on their own contributions. In this case,according to the author, the dependant spouses (usually female) have little or noreturn on their own contributions, which could be interpreted (in that particularcontext) as direct income tax (James, 2009).

Restrictions on accumulation, therefore, can reduce social security expenditure,but may also reduce the participation of beneficiaries (usually women) in the labourmarket; the risk being higher in the case of a prohibition of accumulation withwages rather than with other benefits. In the latter case, this outcome is due to thefact that such a restriction would make it necessary to develop various possiblescenarios in the short, medium and long term, combining complex variables that




36

are surrounded by uncertainties, since evaluating and weighting differentindividuals is difficult. According to James (2009), however, there is the risk that,in a perfect information setting, younger workers would become aware of the factthat older people often receive similar benefits regardless of their contributoryhistory, which could discourage contributions (especially among women).

It is true that the participation rate of women depends on many other factors(social, economic and cultural, for example, which are not necessarily independentof each other). However, the evidence in the literature seems to support thehypothesis that social security rules can themselves have this effect, changingthe behaviour of the insured and their dependants. In other words, even the mostrestrictive measures aimed at rationalizing public spending and preventingundesirable or opportunistic behaviour can create incentives and disincentives thatmake them non-neutral from a labour point of view. The legal framework of awell-designed and calibrated social security system needs to take into account thathuman behaviour is not immutable and can respond to involuntary stimuli.

Reform of survivors’ pensions in Brazil

In 2013, if account is taken of only pension expenditure by the General SocialSecurity Scheme, the special civil service schemes (the various RPPS schemes ofthe Union, the States and Municipalities) and that of the military, total survivors’pension expenditure amounted to about 2.6 per cent of GDP. Data compiled bythe Organisation for Economic Co-operation and Development (OECD) and theStatistical Office of the European Union (EUROSTAT) show that obligationsregarding this type of benefit as a proportion of GDP exceed expectations basedon the demographic structure of Brazil. This is shown in Figures 5 and 6, whereBrazil is compared with 40 other countries.

The dependency ratio, calculated as the ratio of the age group of the populationdefined as economically dependent (in this case, people aged 65 or older) to thepotentially active age group (aged 15–64), reached 9.3 per cent in Brazil in 2013,well below the average (23.5 per cent) and the median (24.5 per cent) for the40 other countries selected for comparison. Only four (10.0 per cent) of thesecountries allocate to survivors’ pensions a proportion of expenditure that is greaterthan 2.0 per cent of GDP. This is in spite of the fact that in these countries about28 per cent of people are older than age 65 (almost triple the rate of Brazil). Thatis, the data indicate that the percentage of Brazilian GDP earmarked for survivors’pensions exceeds that observed for countries with a dependency ratio equal to orhigher than the Brazilian ratio (Figure 6).

The differences in the parameters that determine eligibility for the benefits seemin this context, to a significant extent, to be the fundamental explanation for the




37

pattern of atypical expenditure in Brazil. It seems clear that countries with astronger and more comprehensive social safety net tend to allocate more resourcesin the form of transfers to social security and welfare, and it would hence beinteresting to carry out a cost/benefit analysis to improve comparisons. In Brazilthe cost is high, but the transfers seem to benefit especially people who in theorywould not appear to need subsidized social protection (with a clear mismatchbetween contributions and benefits).

As shown by Ansiliero, Costanzi and Pereira (2014), in a study for 132 countries:(a) 8 per cent of the countries examined imposed some type of minimum periodof contributions; (b) 81.8 per cent applied some condition to ensure the

Figure 5. Expenditure on survivors’ pensions as a percentage of GNP (various yearsand selected countries)

Note: Brazilian expenditure include the entire General Social Security Scheme, all the special schemes of the27 provinces (including the Federal District and the municipalities that have such schemes) and the military.The provinces of Alagoas, Amazonas, Ceará, Maranhão, Pernambuco and São Paulo may include data formilitary police and military firefighters. EUROSTAT data includes total expenditure on survivors’ pensions,including the military. OECD figures (for Japan, Republic of Korea, Canada, Australia, United States, Chile, Israeland Mexico) in turn cover survivors’ pension expenditure, although they may also include administrative costs. Inall cases, only cash transfers paid to dependants (not just widowers and widows, but all types of dependants) areconsidered. In some countries, where survivors’ pensions can be converted into retirement pensions, the data maygenerate some analytical distortions. However, it is assumed here that the comparability of the data is acceptable.

Sources: OECD, Social Expenditure Data Base, SOCX <stats.oecd.org/Index.aspx?datasetcode=SOCX_AGG>;EUROSTAT <ec.europa.eu/eurostat/statisticsexplained/index.php/Social_protection_statistics_-_pension_expenditure_and_pension_beneficiaries>; AEPS and MPS (2013); MPOG-SRH (2010); indicative statisticsresulting from an actuarial evaluation (MPS, 2014). Authors’ interpretation.




38

http://stats.oecd.org/Index.aspx?datasetcode=SOCX_AGG%3e;

http://stats.oecd.org/Index.aspx?datasetcode=SOCX_AGG%3e;

http://ec.europa.eu/eurostat/statisticsexplained/index.php/Social_protection_statistics_-_pension_expenditure_and_pension_beneficiaries


replacement rate reached 100 per cent; (c) 76.5 per cent imposed restrictions onspouses or their equivalent (e.g. minimum age requirement for the spouse orminimum duration of marriage, among other requirements that are notnecessarily cumulative). In Brazil, until 2015 none of these requirements orrestrictions existed in the General Social Security Scheme regulations, and as aresult they were also absent in the regulatory framework for statutory schemesfor public officials.

Comparing this to 23 selected countries in Latin America and the Caribbean(Table 5), it is observed that the great majority apply criteria (normally a set ofcriteria) for the award and maintenance of survivors’ pensions for spouses andassimilated persons, who are generally the main beneficiary group of this type ofbenefit. The most relevant issue, by comparison with Brazil, is the greater rigourapplied in determining the survivors’ pension replacement rate – a differencethat also stands out in a comparison with Europe. A comparison of 32 selectedEuropean countries (Table 5) reveals even more restrictive (sets of) criteria for

Figure 6. Expenditure on survivors’ pensions as a percentage of GDP and thedependency ratio for adults

Sources: OECD, Social Expenditure Data Base, SOCX <stats.oecd.org/Index.aspx?datasetcode=SOCX_AGG>;EUROSTAT <ec.europa.eu/eurostat/statisticsexplained/index.php/Social_protection_statistics_-_pension_expenditure_and_pension_beneficiaries>; AEPS and MPS (2013); MPOG-SRH (2010); indicative statistics resulting from anactuarial evaluation (MPS, 2014). Authors’ interpretation.




39

http://oecd.org/Index.aspx?datasetcode=SOCX_AGG



Table 5. Basic requirements for the award and continuous payment of survivors’pensions – Europe (2014) and Latin America and the Caribbean (2015)

EUROPE: 32 COUNTRIES

Requirement of exclusion period – proportion (%) Yes No Total

100 0 100

Existence of limits and/or conditions (number of dependants) for total replacement rate –proportion (%)

Yes No Total

100 0 100

Maximum value (total) of survivors’ pension varies with number of dependants – proportion (%) Yes No Total

96.9 3.1 100

Maintenance of amount (e.g. non-redistribution of shares to maintain the total value) –proportion (%)

Yes No Total

100 0 100

Penalties – early retirement and/or reduction of benefit amount – proportion (%) Yes No Total

78.1 21.9 100

Maintenance of civil status (cessation of economic activity on marriage or partnership) –proportion (%)

Yes No Total

56.3 43.8 100

Receipt of income from work and/or personal benefit – proportion (%) Yes No Total

56.3 43.8 100

Existence of requirements and restrictions for spouses and assimilated persons

Existence of requirements and restrictions for spouses and assimilated - proportion (%) Yes No Total

84.4 15.6 100

Spouses and assimilated persons – requirements show gender differentiation – proportion (%) Yes No Total

18.8 81.3 100

Minimum age requirement – proportion (%) Yes No Total

68.8 31.3 100

Minimum duration of marriage/partnership – proportion (%) Yes No Total

46.9 53.1 100

Guardianship of minor dependants – proportion (%) Yes No Total

71.9 28.1 100

Requirement of proof of economic dependence – proportion (%) Yes No Total

15.6 84.4 100

Requirement of proof of special needs (disability) – proportion (%) Yes No Total

62.5 37.5 100

LATIN AMERICA AND THE CARIBBEAN: 23 COUNTRIES (EXCLUDING BRAZIL)

Requirement of exclusion period – proportion (%) Yes No Total

100 0 100(Continued)




40

Existence of limits and/or conditions (number of dependants) for total replacement rate –proportion (%)

Yes No Total

91.3 8.7 100

Maximum value (total) of survivors’ pension varies with number of dependants – proportion (%) Yes No Total

91.3 8.7 100

Maintenance of amount (e.g. non-redistribution of shares to maintain the total value) –proportion (%)

Yes No Total

73.9 26.1 100

Penalties – early retirement and/or reduction of benefit amount – proportion (%) Yes No Total

60.9 39.1 100

Maintenance of civil status (cessation of economic activity on marriage or partnership) –proportion (%)

Yes No Total

60.9 39.1 100

Receipt of income from work and/or personal benefit – proportion (%) Yes No Total

4.3 95.7 100

Existence of requirements and restrictions for spouses and assimilated persons

Existence of requirements and restrictions for spouses and assimilated – proportion (%) Yes No Total

87 13 100

Spouses and assimilated persons – requirements show gender differentiation – proportion (%) Yes No Total

39.1 60.9 100

Most frequent requirements and restrictions for spouses and assimilated persons

Minimum age requirement – proportion (%) Yes No Total

56.5 43.5 100

Minimum duration of marriage/partnership – proportion (%) Yes No Total

69.6 30.4 100

Guardianship of minor dependants – proportion (%) Yes No Total

56.5 43.5 100

Requirement of proof of economic dependence – proportion (%) Yes No Total

56.5 43.5 100

Requirement of proof of special needs (disability) – proportion (%) Yes No Total

56.5 43.5 100

Note: Conditions are in many cases combined and may be used only as attenuating considerations for specificrequirements. Latin America and the Caribbean: Argentina, Barbados, Bermuda, Bolivia, Chile, Colombia, CostaRica, Cuba, Dominica, Dominican Republic, Ecuador, Guatemala, Guyana, Honduras, Jamaica, Mexico,Nicaragua, Panama, Paraguay, Peru, Trinidad and Tobago, Uruguay, Bolivarian Republic of Venezuela. Europe:Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Estonia, Finland, France, Germany, Greece,Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal,Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey, United Kingdom.

Sources: ISSA and SSA (2014, 2015).

Table 5. Basic requirements for the award and continuous payment of survivors’pensions – Europe (2014) and Latin America and the Caribbean (2015) - Continued




41

survivors’ pensions, the most important difference being (even in the post-reformperiod of 2014–15) that the assessment of the monetary value of the benefit ismade at the time of its award (the level being related to the number of personsdependant on the insured person) and during its continued payment (withoutany redistribution of shares forfeited by one beneficiary among the remainingbeneficiaries).

In Brazil, the lack of a minimum period of contributions and minimumduration of marriage made possible opportunistic behaviour by insured persons.It was fully (and legally) possible, for example, that a single monthly contribution– in the case of self-employed persons – could generate a lifelong survivors’ pensionat the maximum General Social Security Scheme ceiling. The pension could also begranted on a lifelong basis to spouses without any conditionality – to young widowsand widowers without children, with full capacity to work and even where theywere employed with a decent level of income.

As already stated, the most notable difference is the definition of the maximumamount of benefit for a group of pensioners dependent on the same insured personwho has died. In most pension schemes, the final amount of the benefit is the sumof the shares allocated to the dependants, which very often cannot be changed withthe death of a dependant or the loss of rights by any of them. Therefore, in manycases no guarantee is given of a replacement rate of 100 per cent of the amount thatthe deceased would have been entitled to if he or she had eventually been grantedsome other personal benefit (reference value). In Brazil there was (and still is) aguarantee for dependants of 100 per cent of the pension that the beneficiary wasreceiving or would have been entitled to receive if they had retired on grounds ofdisability, and the termination of one dependant’s share leads to redistributionof the total value of that person’s share among the other dependants.

In Brazil before the reforms of 2014–15, under the General Social SecurityScheme the only real limitation concerned the payment of more than onesurvivors’ pension to a beneficiary, without prejudice to the right to choose themost advantageous benefit and combine it with other types of pension benefits(such as an old-age pension, for example). Taking as a reference the countries ofthe European Union (the group for which there are comparable data of betterquality), it may be observed that in Brazil the incidence of the accumulation ofbenefits (old-age and survivors’ pensions, as in the data gathered by the IBGENational Household Surveys (various years)) is also at variance with more usualinternational practice.

Although the European figures (Table 6), like those for Brazil, also suffer fromsome methodological limitations, it may be observed that the accumulation ofbenefits is more frequent in Brazil (33 per cent in 2014): only one of the 33 selectedcountries (Hungary) has more than 25 per cent of its beneficiaries receiving morethan one type of benefit (old-age pension and/or survivors’ pension under any of




42

the existing pension schemes); 33 per cent of countries show no accumulation; and60 per cent of them score a maximum accumulation percentage of 10 per cent.

On this point, Brazil follows the Latin American trend whereby there is nosanction for the accumulated receipt of pension benefits: of the 23 countriesstudied, only Peru imposes a limit, and only for the accumulation of survivors’pensions with income from work by dependent widowers – the requirement ofeconomic dependency applies only to those aged 60 or older at the time of deathof the insured person. In Europe, approximately 38 per cent of countries limit insome way the accumulation of survivors’ pensions with income from work;47 per cent prohibit the accumulation of this type of benefit with personal benefits(retirement) or at least establish a cap on the sum of their amounts; and 56 per centapply at least one such restriction.

International comparison reinforces the perception that differences in theframeworks governing such benefits may be associated with greater rigour inEurope regarding the rules on the accumulation of benefits and also the rulesin force for the award and maintenance of survivors’ pensions. Where eligibility(recognition of such entitlement) is subject to more conditions, excesses arenaturally contained and the prohibition of accumulation is less relevant.

In any case, the accumulation of survivors’ pensions with other income, and inparticular with other social security transfers, is not necessarily serious, since theytend to be associated with different funding sources. In other words, the problemis not so much accumulation as the combination of this right with over-generous

Table 6. Proportion of beneficiaries receiving more than one benefit (survivors’pensions or other welfare benefits) – various countries (2012)

Countries (33) Percentage ofaccumulation

Proportion ofcountries (%)

Estonia, Croatia, Iceland, Latvia, Malta, Netherlands,Norway, Poland, Romania, Serbia, United Kingdom

0 33

Bulgaria, Ireland, Turkey 0–5 9

Cyprus, Denmark, Greece, Luxembourg, Slovenia, Spain, 5–10 18

Austria, Sweden 10–15 6

Finland, Germany, Italy, Portugal, Switzerland 15–20 15

Czech Republic, France, Lithuania, Slovakia 20–25 12

Hungary 25< = accumulation 3

Belgium Information unavailable 3

Total 100

Source: EUROSTAT <ec.europa.eu/eurostat/data/database>. Author’s interpretation.




43

http://ec.europa.eu/eurostat/data/database

rules for the award and continuous payment of the benefit in Brazil, which hasgenerated a risky situation permitting the payment of many benefits for ratherlong periods of time.

Because of the issues raised, as well as the pattern of fiscal adjustment thatprevailed after the counter-cyclical measures demanded by the 2008–09 crisis andthe significant worsening of Brazil’s public accounts since 2014, an attempt wasmade to launch a comprehensive review of the rules for survivors’ pensions inlate 2014, through Provisional Measure (PM) 664 of 30 December. However, thefinal outcome of the reform was significantly watered down by the BrazilianCongress of Deputies and was applied by Act No. 13,135 of 17 June 2015.

The promulgation of Provisional Measure 664/2014 was a clear attempt to bringthe General Social Security Scheme more into line with the standards applied at theinternational level. The rules applicable before 664/2014, as well as the changesproposed in the Provisional Measure and those finally approved are set out in Table7. In the case of survivors’ pensions, the measures proposed in PM 664 were as follows:(a) introduction of an exclusion period of two years of contributions, sincepreviously there was only the requirement of being an insured person, such thatit was possible to obtain lifelong benefit at the maximum level after only a singlecontribution;(b) the introduction of a minimum duration of marriage or partnership of twoyears for the spouses, a condition that did not exist previously;(c) the introduction of conditions for the award of life annuities, which werepreviously awarded to all spouses without conditions, even to young spouseswithout children and with other sources of income. The government’s proposalwas that spouses would only be entitled to a life annuity with a life expectancy, atthe time of the partner’s death, of up to 35 years; spouses with a life expectancy ofmore than 35 years (younger people) would receive a pension for a periodof between 3 and 15 years, such that the younger the beneficiary, the shorter theduration of the benefit;(d) the end of the right to survivors’ pensions for dependants convicted of a crimethat resulted in the death of the insured person;(e) changes in the value of the calculated benefit that would no longer equal 100 percent of the retirement pension that the insured person was receiving or would havebeen entitled to, but reduced to 50 per cent thereof plus 10 per cent for eachdependant, and subject to the minimum wage limit;(f) the end of so-called share redistribution, a system whereby in the event of loss ofdependency status by one of the dependants (for example, when a child reaches age21), the value of his or her share is redistributed among the remaining dependants.

The main objective of the changes to the rules governing survivors’ pensionsconcerned the General Social Security Scheme, but most of the changes also coveredthe pension scheme for federal public officials, except for the change in the




44

Table 7. Rules on survivors’ pensions in Brazil – before PM 664/2014, as proposed inPM 664/2014, and as adopted (Act No. 13,135/2015)

Criterion Rule before PM 664 Rule proposed inPM 664/2014

Rule as approved byAct 13,135/2015

Exclusionperiod

Award without reference to anyexclusion period, the onlycondition being the status ofinsured person.

Requirement of an exclusionperiod of two years ofcontributions, except in thecase of employment accidentsor employment or occupationaldisease; this also coveredfederal public officials.

The exclusion period wasreduced to 18 months ofcontributions, with however theintroduction of the payment offour months’ benefit for insuredpersons not meeting thiscondition; there was also achange in the exceptionsconcerning accidents of anynature and employment oroccupational illnesses. Suchchanges were also extended tocover federal public officials.

Minimumduration ofmarriage orstablepartnership

No requirement concerningany minimum duration.

Introduction of a requirement oftwo years of marriage or stablepartnership. Exceptions wherethe death was due to anaccident occurring aftermarriage or where the spouseis incapable of performing paidactivity guaranteeing a meansof subsistence. This changewas also to be extended tocover federal public officials.

The minimum period of twoyears of marriage or stablepartnership was retained, withhowever the introduction of thepayment of four months’ benefitfor insured persons not meetingthis condition. This requirementdoes not apply where the deathresulted from an accident orany type of employment oroccupational illness, nor tospouses with disabilities. Thischange was also extended tocover federal public officials.

Duration ofspouse’sbenefits

Lifelong survivors’ pensionregardless of the age of thespouse (widow/widower) at thetime of death, even wherechildless, with full capacity forwork and without any need toprove economic dependency,which was presumed.

The duration should depend onthe life expectancy of thesurvivor at the time of thespouse’s death. Where suchlife expectancy is of up to35 years (aged 44 or older) thebenefit is lifelong; durationvaries from 3 to 15 years forthose with a life expectancy ofover 35 years. To be revisedannually. This would alsoextend to cover federal publicofficials.

Duration defined based on theage of the survivor at the timeof the spouse’s death, beinglifelong for a spouse (widow/widower) of 44 years or older,and from 3 to 20 years for thoseyounger than age 44. Ages tobe revised at least once everythree years or where the lifeexpectancy indicator increasesby one year. This change wasalso extended to cover federalpublic officials.

Homicide ofthe insuredperson by adependant

No legal restrictions on thepayment of a survivors’pension to a dependant whocommitted a crime resulting inthe death of the insuredperson.

Pension entitlement excludedfor the dependant committing acrime resulting in the death ofthe insured person. This wouldalso apply to federal publicofficials.

Loss of pension entitlementretained for a dependant foundguilty of an intentional crimeresulting in the death of theinsured person, but only afterfinal sentence is pronouncedby the court. This change wasalso extended to cover federalpublic officials.

(Continued)




45

calculation of the pension value.While the requirement of aminimumof two years ofmarriage or partnership was retained and aminimum of 18months of contributionshas been introduced, these measures have been offset by creating a payment of4 months of benefit for the spouses of insured persons who do not meet theseconditions.

Although the final version has introduced conditionalities on lifelongpensions for spouses, the conditions and the duration of the benefit have beenrelaxed regarding the age of the widow or widower at the time of the insuredpartner’s death. As from age 44 the pension is lifelong, and for those youngerthan age 44 the duration varies between 3 and 20 years (and not from 3 to15 years as proposed in PM 664); this must be revised at least once everythree years as regards life expectancy at birth (or where the life expectancyindicator increases by at least one year) and no longer annually, as proposedby PM 664.

The changes in the amount of the pension, in particular the reduction from100 per cent to 50 per cent of the retirement pension for each dependant andthe proposed abolition of the share redistribution system were rejected byCongress, and there were hence no changes in the benefit amount. These were,without a doubt, the harshest measures, and therefore those that would have hadthe greatest fiscal impact. In contrast, approval was given for the loss of entitlementto a survivors’ pension for a dependant found guilty of a crime resulting inthe death of the insured person, but with one change: while PM 664 linkedloss of entitlement to the conviction of the dependant (as in the first examples),the final version makes loss of entitlement apply only after final sentence ispassed.

Table 7. Rules on survivors’ pensions in Brazil – before PM 664/2014, as proposed inPM 664/2014, and as adopted (Act No. 13,135/2015) - Continued

Criterion Rule before PM 664 Rule proposed inPM 664/2014

Rule as approved byAct 13,135/2015

Amount ofbenefit

100% of the retirement benefitthat the deceased spouse wasreceiving or would have beenentitled to receive if retiring ongrounds of disability. Whereone dependant losesentitlement, the value of theirbenefit is shared among theother beneficiaries(redistribution of shares).

Value to be reduced to 50%,plus 10% for each dependant,of the amount that thedeceased spouse wasreceiving or would have beenentitled to receive if retiring ongrounds of disability, withretention of the guarantee ofmeeting the minimum wagerequirement. Abolition of theredistribution of shares.

Rules proposed in PM 664were rejected, and theprevious rules were reinstated.

Source: Authors’ interpretation, based on Provisional Measure (PM) No. 664 of 30 December 2014, and ActNo. 13,135 of 17 June 2015.




46

Closing remarks

As shown in the first section of this article, greater participation by women inthe labour market has led to a greater accumulation of retirement andsurvivors’ pensions by survivors in Brazil. In addition, the previous ruleguaranteeing lifelong pensions to spouses without any conditionality, even toyounger people without children and with full capacity to work or evenemployed with high incomes, ended up creating distortions in the socialprotection system by guaranteeing incomes for people who already have othersources of income.

In this context, the presumption of economic dependency for spouses must bere-evaluated. Moreover, as shown from an econometric perspective, this rule hashad a negative impact on the participation of women in the labour market.Although the gap is partly explained by discrimination, the award of a lifelongpension without any conditionality also seems to have influenced the differentlevels of labour market participation for men and women.

Brazil still has a very peculiar combination of survivors’ pension rules. Forexample, benefit has been granted purely on the basis of the insured person’s status,without any need to have satisfied a minimum period of contributions or durationof marriage, for life and without any conditionality, which allowed opportunisticbehaviour. As a result of these peculiar rules, Brazil has a pattern of spending onsurvivors’ pensions as a percentage of GDP that is much higher than expectedgiven its current demographic structure, such expenditure reaching almost3 per cent of GDP. Although the 2014–15 reform has made it possible to bringthe system more into line with international rules, the proposal was significantlytoned down by the Brazilian Congress.

In general, it should be clear from the case under review that there is a need toacknowledge that social security must adapt to changes in the labour market, andthat social security rules can have a significant impact on the world of work.

Bibliography

AEPS;MPS. 2013. Anuário estatístico da previdência social (Infologo). Brasilia, Ministério da

Prevedência Social.

Ansiliero, G.; Costanzi, R. N.; Pereira, E. da Silva. 2014. “A pensão por morte no

âmbito do regime geral de previdência social”, in Revista Planejamento e Políticas

Públicas, No. 42.

Barbosa, A. L. Neves de Holanda; Corseuil, C. H. Leite. 2013. Bolsa Família, occupational

choice and informality in Brazil (Working paper, No. 118). Brasilia, International Policy

Centre for Inclusive Growth.




47

http://www3.dataprev.gov.br/infologo/

http://www.ipea.gov.br/ppp/index.php/PPP/article/view/271

http://www.ipea.gov.br/ppp/index.php/PPP/article/view/271

http://www.ipc-undp.org/pub/IPCWorkingPaper118.pdf

http://www.ipc-undp.org/pub/IPCWorkingPaper118.pdf

Decoster, A.; Orsini, K.; Van Camp, G. 2007. Labour market responses of survival pensioners:

Estimating a labour supply model and predicting the effect of the reform. Leuven, CES KU

Leuven.

ECLAC; ILO. 2014. The employment situation in Latin America and the Caribbean –

Conditional transfer programmes and the labour market. Santiago, Economic Commission

for Latin America and the Caribbean, International Labour Organization – Subregional

Office for the South Cone of Latin America.

Foguel, M. N.; Barros, R. Paes de. 2010. “The effects of conditional cash transfer

programmes on adult labour supply: An empirical analysis using a time-series-cross-

section sample of Brazilian municipalities”, in Estudos Econômicos, Vol. 40, No. 2.

Gruber, J.; Orszag, P. 2003. “Does the social security earnings test affect labor supply and

benefits receipt?”, in National Tax Journal, Vol. 56, No. 4.

IBGE. 1992. Pesquisa Nacional por Amostra de Domicílios: Microdados 1992. Brasilia,

Instituto Brasileiro de Geografia e Estatística.





ISSA; SSA. 2014. Social security programs throughout the world: Europe, 2014. Washington,

DC, Social Security Administration.

ISSA; SSA. 2015. Social security programs throughout the world: The Americas, 2015.

Washington, DC, Social Security Administration.

James, E. 2009. Rethinking survivor benefits (Social protection discussion paper, No. 0928).

Washington, DC, World Bank.

Knapp, D. 2014. The effect of social security auxiliary spouse and survivor benefits on the

household retirement decision (16th Annual Joint Meeting of the Retirement Research

Consortium, Washington, DC, 7–8 August). Ann Arbor, MI, University of Michigan.

MPOG-SRH. 2010. Boletim estatístico de pessoal, Vol. 15, No. 167.

MPS. 2014. Quadro de demonstrativo de resultado da avaliação atuarial (DRAA) do exercício

de 2014 (Ano base 2013). Brasilia, Ministério da Previdência Social.

Nishiyama, S. 2015. The joint labor supply decision of married couples and the social security

pension system (Working paper, No. 17). Lancaster, Lancaster University – Department of

Economics.




48

http://dx.doi.org/10.2139/ssrn.1021160

http://dx.doi.org/10.2139/ssrn.1021160

http://www.ilo.org/wcmsp5/groups/public/---americas/---ro-lima/documents/publication/wcms_244281.pdf

http://www.ilo.org/wcmsp5/groups/public/---americas/---ro-lima/documents/publication/wcms_244281.pdf

http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0101-41612010000200001



http://economics.mit.edu/files/6438

http://economics.mit.edu/files/6438

http://www.ibge.gov.br/home/estatistica/populacao/trabalhoerendimento/pnad2013/microdados.shtm



http://documents.worldbank.org/curated/en/500701468331226085/pdf/529190NWP0Box345585B01PUBLIC100928.pdf

http://crr.bc.edu/wp-content/uploads/2014/06/Panel-3_3-Knapp1.pdf

http://crr.bc.edu/wp-content/uploads/2014/06/Panel-3_3-Knapp1.pdf

http://www.planejamento.gov.br/secretarias/upload/Arquivos/servidor/publicacoes/boletim_estatistico_pessoal/2010/bol167-mar2010.pdf

http://www1.previdencia.gov.br/sps/app/draa/draa_default.asp

http://www1.previdencia.gov.br/sps/app/draa/draa_default.asp

http://www.research.lancs.ac.uk/portal/en/publications/the-joint-labor-supply-decision-of-married-couples-and-the-social-security-pension-system(99de5984-0b10-4f0e-bd2e-31c9d54a8ef5).html

http://www.research.lancs.ac.uk/portal/en/publications/the-joint-labor-supply-decision-of-married-couples-and-the-social-security-pension-system(99de5984-0b10-4f0e-bd2e-31c9d54a8ef5).html

WORLDVIEW

Worldview: An introductionto social security in Georgia

This Worldview section presents two articles that address different policychallenges confronting the social security system of Georgia. As one of the formerSoviet Republics, Georgia remains relatively under‐reported in the wider socialprotection literature and its social security system is little‐known. One importantsource of empirical information concerning social security provisions in Georgiais found in the International Social Security Association (ISSA) Country Profiledatabase, the national profile for which has been developed by the ISSA inpartnership with the US Social Security Administration as part of joint work toproduce the key reference publication Social Security Programs throughout theWorld. Prior to this issue, the Review had never published a single analytic articlededicated to Georgia. While this first issue of 2017 was being planned, serendipitydetermined that the Social Services Agency of Georgia, the country’s national socialsecurity administrative body, would become an Affiliate member institution of theISSA, the Review’s sponsor. Given that an important objective of the ISSA is toprovide knowledge to its member institutions, the following two articles are indeedtimely and appropriate.

In thefirst article,DimitriGugushvili uses awide lens to examine the chronologicaldevelopment of social security in Georgia prior to and during the Soviet era, and thento map the recent evolution of the country’s social protection system sinceindependence in 1991. The author traces this trajectory through what was a difficultfirst decade of transition and thereafter into a period that has witnessed importanteconomic growth and the modernization of public administration. Significant inthis regard has been the modernization of the administration of social securityand health care programmes. Gugushvili discusses the recent expansion of Georgia’ssocial protection system and suggests optimism for the future, in spite of a numberof important challenges – especially in relation to the evidence of the system’slimited effect on reducing widespread poverty and the absence of unemploymentprotection. The longer‐term objective, in the author’s view, must be for Georgiato construct a redistributive social protection system, a welfare state, which issupportive of inclusive growth.

The second article addresses the challenge of pension system reform inGeorgia, with a specific focus on the important role played by non‐contributorypensions in tackling elderly poverty. Tamila Nutsubidze and KhatunaNutsubidze report that Georgia’s national social security system has almostuniversal non‐contributory basic pension coverage. Their analysis sets out that



49

the non‐contributory basic pension has proven effective in mitigating povertyamong pensioners, especially those living alone or in small households. Fiscalconstraints and population ageing, however, now raise questions regarding thefuture adequacy of the programme and its sustainability. In response to thesechallenges, and in the context of policy proposals to introduce a definedcontribution pension system, the authors present a number of pension systemreform scenarios, highlighting the difficult trade‐offs that are implicit in each.Using Georgian experience as a case study, it is argued that the country’sbasic pension model holds value for other low‐ and middle‐income countrieslooking to enhance old‐age coverage and better mitigate poverty.

Roddy McKinnonEditor

Worldview: An introduction to social security in Georgia



50

WORLDVIEW

The evolution of the Georgiansocial protection system sinceindependence: An unclear

future despite recent expansionDimitri Gugushvili

Catholic University of Leuven, Belgium

Abstract In the first decade of transition, the Georgian socialprotection system experienced a major retrenchment as thegovernment struggled to finance welfare provision in the faceof massive economic contraction and the near collapse ofpublic institutions. Since 2004, this trend has been reversed,with the economy returning to a fast growth path and publicadministration improving considerably. Recent reforms,including the notable introduction of universal public healthinsurance, are welcome steps towards building a modernwelfare state. Major challenges still remain, however,especially in relation to the system’s limited effect onwidespread poverty. Decelerating growth, the lack of strongpro-welfare actors, and the absence of positive external pullfactors may stall or prevent future growth, but the changingnature of the social contract between the people andgovernment, as well as Georgian politicians’ growingrecognition of the importance of the welfare system forinclusive growth, leaves ample space for optimism.

Keywords social security administration, social securityplanning, welfare state, Georgia, Europe

Address for correspondence: Dimitri Gugushvili, Centre for Sociological Research, KU Leuven,Parkstraat 45, Box 3601, 3000 Leuven, Belgium; email: [email protected].

The author thanks the two anonymous referees for their constructive comments and suggestions.



51

Introduction

Despite its relatively short history, the Georgian social protection system hasexperienced major changes in the past 25 years. As in other former SovietRepublics, the first years of transition saw a major cutback of the inheritedextensive welfare system, though the retrenchment in Georgia was far more severethan in other countries due to the large scale of economic contraction and the nearcollapse of public institutions. For many years formal welfare entitlements existed,by and large, only on paper. The system was reanimated and has continued toexpand under two successive governments since 2003. Nevertheless, to meet theminimum objectives of a modern welfare system, far more substantive changesare required which, in turn, will require the emergence of strong pro-welfarepolitical actors.

This article provides a comprehensive analysis of the Georgian social protectionsystem, thereby contributing to the knowledge base on welfare systems in transitioncountries. It is organized as follows. The first section briefly describes the Sovietwelfare system and documents three major phases in the evolution of theGeorgian social protection system since 1991. The second section maps the maincomponents of the existing system. The third section identifies its key strengthsand weaknesses. The final section discusses the factors that may hinder orfacilitate its further development in the coming years.

Evolution of the Georgian social protection system

Soviet legacy

The concept of formal social protection in Georgia was introduced by the SocialDemocratic government in 1921. The Parliament of the first Democratic Republicof Georgia adopted what was one of the most progressive Constitutions of its time.It guaranteed the right to primary education, social insurance and social assistancebenefits, a minimum wage, maximum working hours, maternity pay andprotection and special protection measures for vulnerable workers (i.e. pregnantwomen and children aged 16–18), as well as the right to form trade unions. Italso asserted that the protection of small producers (farmers and artisans) fromexploitation was the “state’s special responsibility” (Democratic Republic ofGeorgia, 1921, Article 116), and required the establishment of an employmentagency and labour inspectorate. The 1921 Constitution was adopted two weeksprior to the outbreak of the war with the Russian Federation.

The Social Democratic government, having been forced to flee the countryshortly after their military defeat, had neither the time nor the capacity to

The evolution of the Georgian social protection system



52

implement any of these measures. Therefore, formal social protection mechanismswere introduced later in Georgia, during the Soviet period. Albeit distinctive, theSoviet welfare system in some ways resembled the welfare states of advancedcapitalist countries (Gugushvili and Sundberg, 2012; Cook, 2010; Chandler, 2001;Fajth, 1999). It was constructed around three main pillars: (i) guaranteedemployment; (ii) cash transfers (mainly old-age and disability pensions) andoccupational in-kind benefits; and (iii) free and universal education and healthcare, and also some residential services for older persons and for children deprivedof parental care (Orenstein, 2008; Standing, 1996).

In addition to social protection mechanisms, Soviet citizens benefited fromhighly subsidized food, transport, energy and other basic consumer goods(Orenstein, 2008; Deacon, 2000; Standing, 1996), which enhanced the purchasingpower of wages which were rather low by OECD standards. In line with theoverall political system, the decision-making and financing of welfareprogrammes were highly centralized. In this system, the delivery of somebenefits, especially those that were in-kind, was delegated to trade unions andlarge enterprises, which effectively served as extensions of local authorities(Standing, 1996). Total social spending was estimated to be in the range of15–20 per cent (Aidukaite, 2004), which until the 1970s exceeded the socialspending of an average OECD welfare state (Fajth, 1999).

The Soviet welfare regime had a number of obvious strengths (Orenstein, 2008;Deacon, 2000; Standing, 1996; Deacon, 1992 and 1993). The vast majority of theworking-age population was employed in full-time jobs and all retirees wereentitled to different types of pensions (Orenstein, 2008; Fajth, 1999). Workersenjoyed strong employment rights and female employees benefited fromgenerous maternity leave and childcare services. In addition to salaries, mostemployees were entitled to a wide range of occupational benefits. Differences inwages between managers and manual workers were relatively small (Deacon,2000). Selected groups of workers, for example miners and construction workers,even had higher salaries than middle-class professionals, such as doctors andteachers (Standing, 1996). Illiteracy and homelessness had been fully eradicated.Free and universal access to all forms of health care and education helped toenhance human capital. Despite hidden privileges for the ruling elite, overallinequality was low by OECD standards (Milanovic, 1995).

With closer inspection, each of these achievements turned out to be much lessimpressive (Deacon, 1992 and 2000; Manning, 1992). Full employment was achievedat the expense of low productivity and latent unemployment/underemployment.Despite extensive workplace safety regulations, some workers were frequentlyexposed to hazardous and disabling environments. Employment-related benefits,especially old-age pensions, perpetuated existing lifetime income inequalitiesbetween the high- and low-paid workers, and hindered labour mobility (Standing,




53

1996; Manning, 1992). In addition, large groups of workers, especially thoseemployed in collective farms, had very low pensions (Aidukaite, 2004; Atkinsonand Micklewright, 1992). Moreover, discretionary indexation of benefits meantthat their real value was often reduced by inflation (Deacon, 2000). The quality ofhousing was often poor, and overcrowded accommodation was commonplace(Deacon, 1993; Manning, 1992). Access to health care services was not entirelyfree; unofficial out-of-pocket payments to doctors and nurses was the rule ratherthan the exception. The overall provision of health services was highly inefficientand the quality varied between the centre and peripheries (Deacon, 1992).

Underlying the whole system was a regressive distribution of benefits. Those inprivileged positions (usually connected to the ruling elite) received the best qualityhousing, medical services and consumer goods, which were in short supply, such asautomobiles (Deacon, 2000; Atkinson and Micklewright, 1992). Besides, the Sovietwelfare state did not succeed in eradicating absolute poverty. Drawing on officialdata, Atkinson and Micklewright (1992, pp. 239–245) estimate that 11 per centof Soviet citizens were “less provided” for, a Soviet euphemism for povertymeasured against the threshold of a normative minimum basket of goods andservices. Inequalities within different parts of the Soviet Union must also havebeen substantial, as the incidence of “less provided” families was 51 per centin Tajikistan compared to 1.9 per cent in Estonia. Georgia’s poverty rate(13 per cent) was slightly higher than the Soviet Union average.

The retrenchment phase (1991–2003)

The Soviet welfare system was tailored to its particular type of centrally plannedeconomy, which was impossible to maintain intact after the economic modelcollapsed. The welfare system of the newly-independent Georgia came underhuge strain due to the near collapse of the state and the catastrophic fall in publicrevenues, especially in the first years of transition – between 1990 and 1994economic output fell by three-quarters (Falkingham, 2003). While some of themajor commitments of the Soviet welfare system, such as guaranteed employment,housing provision and various subsidies, were immediately abandoned, PresidentShevardnadze’s government formally maintained other core programmes, such asnumerous types of pensions and free services. The right to universal andcomprehensive health care was revoked later, in 1996, but the government formallymaintained the commitment to provide a universal basic package of primaryservices and some types of medical treatment for specific vulnerable groups(OSGF, 2012).

Despite this significant retrenchment, Shevardnadze’s government struggled tofind the means to finance the remaining social programmes, as well as the state’s




54

other basic responsibilities, such as public servants’ salaries. To further reducesocial spending, the retirement age was raised (from age 55 to age 60 for women;from age 60 to age 65 for men) and the real value of social transfers was allowedto erode sharply, initially owing to rampant inflation and then by means ofinadequate indexation. Likewise, spending on services was cut several-fold, buteven after these measures the government was not able to ensure the timely payout of transfers or salaries to public service providers. For example, in 2000,planned public health expenditure was set at 0.6 per cent of GDP (correspondingto USD 11 per person a year), however by the end of that year the governmenthad not managed to fully allocate even this minimum amount (Gamkrelidzeet al., 2002).

There were also some attempts to adjust the welfare system to new risks. Similarto other transition countries, Shevardnadze’s government introduced several newsocial transfers: unemployment allowance and categorical1 social assistancebenefits; namely, Internally Displaced Persons (IDP) Allowance and FamilyAssistance (OSGF, 2012). It also established the State Employment Agency.However, the actual coverage of those benefits, with the exception of the IDPAllowance, was very small, the level of transfers was very low compared tosubsistence minimum levels, and the payments were frequently delayed. Forinstance, in 2000, the government allocated 4.9 million Georgian Lari (GEL)for unemployment allowance, whereas the number of unemployed was estimatedto exceed 700,000 (Tvalchrelidze, 2003). Therefore, only three per cent of theunemployed actually registered for the benefit, but the payment of the GEL 20 amonth benefit (approx. USD 10) remained irregular. Likewise, due to a lack ofadequate financial and administrative resources, the State Employment Agencyexisted, by and large, on paper alone (Tvalchrelidze, 2003). Similarly, while thestrict Soviet labour market regulations were maintained, the government madeno serious effort to enforce them.

The near collapse of the welfare system, combined with a sharp decline inoutput, resulted in a dramatic worsening of living conditions for the majority ofGeorgia’s citizens. Poverty and unemployment reached unprecedented levels and,for many families, labour migration and the income it provided became the onlysource of livelihood.

Reanimation of the system (2004–2012)

If the first decade of transition can be summed up as “lost” due to a sharp fall ineconomic output followed by years of stagnation, the 2000s proved to be one of

1. Categorical implies benefits that are targeted towards specific groups, such as disabled people,children or orphans, and there is no requirement for means testing.




55

sweeping changes. Shortly after winning the presidential election with around95 per cent of votes, President Saakashvili and his United National Movement(UNM) government started to implement a comprehensive package of radicalneoliberal reforms, including all-out privatization, full deregulation, reduction oftaxes, downsizing of the public sector, and trade liberalization. At the same time,major efforts were made to curb corruption and reduce the size of the informaleconomy (World Bank, 2012). The efforts to increase public revenues, byenforcing compliance, were particularly successful – between 2003 and 2012 thestate budget as a share of GDP nearly doubled, while it increased more thanseven-fold in nominal terms (Geostat, 2016). Due to the UNM government’sideological stance, improving the welfare system was not a central concern, butnonetheless it implemented a number of measures that were critical to reanimatethe system.

The first step was to clear the pension arrears inherited from the previousregime. This was followed by abolishing the minimum contributionrequirements for old-age pensions and merging 84 different types of pensionsinto four categories – Old-age, Disability, Survivors and Victims of Sovietrepressions. Privileged pensions were maintained for certain categories of formerpublic servants (judges, members of parliament, military and police personnel,etc.), but these pensions were renamed as “Compensations”. The hand-to-handmethod of benefit delivery was replaced by electronic transfers through thebanking system. Instead of earmarked social contributions, pensions started to befinanced from general taxation. Furthermore, from 2007 to 2012, old-agepensioners received small pension supplements based on the number of yearsworked – the maximum supplement was GEL 10 a month for those with anemployment record of 25 years or more. Most importantly, the governmentcontinuously increased all pensions, though at different rates. The main emphasiswas placed on Old-Age Pensions, which were gradually raised from GEL 14 amonth (approx. USD 7) in 2003 to GEL 125 a month (approx. USD 75) in 2012.2

The main systemic change included the development of a means-testingmechanism and the introduction of targeted social assistance. Prior to this, thetargeting of social assistance was based on categories, such as single pensioners,IDPs, orphans, and families with many children. Also, some of these categoricalbenefits were in-kind. In addition, there was an unemployment allowance, but itscoverage was extremely low (Tvalchrelidze, 2003). In 2005, the Ministry ofLabour, Health and Social Affairs (MOLHSA) set up a database of sociallyvulnerable families and introduced a requirement for social assistance applicantsto undergo a proxy means-testing procedure. At the same time, the

2. The dollar value of transfers fluctuates across years not only because of indexation, but due tochanges in the exchange rate.




56

unemployment allowance was abolished. Several categorical benefits weremaintained, but were monetized and closed for new entrants. The means-testingmechanism was later also used for targeting free health insurance and variousone-off benefits.

The UNM government also undertook a major reform aimed at thede-institutionalization of children placed in state care. With technical andfinancial support from donor3 and civil society organizations,4 it helped developand professionalize the social work profession and gradually replaced largeorphanages with alternative forms of care, such as foster care, small group homesand day care services (O’Brien and Chanturidze, 2009), while also putting anumber of mechanisms in place to minimize the number of children placed instate care.

The UNM’s labour market and health care reforms across 2004–2012 were fullyin line with textbook neoliberal prescriptions. Despite strong resistance from tradeunions, as part of the radical deregulation agenda the government adopted a newlabour code in 2006 that effectively minimized the rights of employees by givingemployers full freedom in hiring and firing (Muller, 2012; Shvelidze, 2012;International Trade Union Confederation, 2008). In the case of dismissal,employers’ responsibilities extended only to providing severance pay equal to onemonth’s salary; but only if the employee had worked for the employer for morethan six months. Limitations on the use of fixed-term contracts were removed.The minimum wage was abolished and although the number of working hourswas set at 41 hours, it could be extended if agreed by both parties. The rulesconcerning overtime pay were abolished and occupational health and safetyregulations became limited to a small list of hazardous occupations. Applying theOECD methodology for calculating the rigour of employment protection, Muller(2012) estimates that after the adoption of the new labour code, Georgia’s scoreon this index decreased from 1.9 to 0.5 (out of possible maximum of 6),compared to an average score of 2.2 for OECD countries.

Similarly, as part of the wider privatization agenda, the UNM governmentprivatized the bulk of health care infrastructure and confirmed it would abandonthe commitment to universal health care provision (Government of Georgia,2007; WHO, 2009). Instead, it began to purchase comprehensive health insurancepackages for low-income families (identified through proxy means-testing), whilealso funding several disease-specific and public health programmes. Public healthexpenditure grew considerably in nominal terms, but as a share of GDP it stayedone of the lowest in Europe, fluctuating between 1.2 to 2.3 per cent of GDP(World Bank, 2016).

3. Notably, UNICEF, the EU Mission and USAID.4. Notably, EveryChild, Save the Children, and World Vision.




57

Continued expansion (2013–2016)

In 2012, the UNM Party lost the parliamentary elections and a coalition consistingof six parties (Georgian Dream Party, Republican Party, Free Democrats, NationalForum, Industrialists, and Conservative Party) formed the new government. Socialpolicy issues clearly were not the coalition’s top priority, with only vaguecommitments to improving welfare provision occupying a page and a half in the72-page election manifesto (Georgian Dream, 2012). Overall, the programmeidentified four priority areas:• Ensuring universal access to health care through extending the previouslymeans-tested health insurance to the whole population, while also maintainingstate-funded health care programmes.• Introducing mandatory private pension insurance and gradually limitinggovernment responsibilities to the regulation of private pension funds and theprovision of the state social pension and/or top-ups for low-earners or thosewith a limited employment record. At the same time, a pledge was made to raiseall existing pensions to the level of a subsistence minimum (GEL 149 a monthfor working-age males in December 2012).• Expansion of means-tested social assistance, with the amount of the transfercalculated based on the level of need.• Decentralization of social service provision while maintaining centralizedfunding.

In addition, the manifesto pledged to revise the labour code, which it describedas “tailored only to the interests of employers”, so as to ensure “strict observance ofthe core rights of employees” (Georgian Dream, 2012; author’s translation).

At the time of ending its term of office in October 2016, the coalitiongovernment had made considerable progress in each of these areas, except forthe decentralization of social service provision. The main achievement has beenthe extension of the basic health insurance scheme to the whole population – theshare of people with public health insurance jumped from 23 per cent in 2011 to87 per cent in 2015 (Mshvidobadze, 2016; Huby et al., 2012).

The coalition government also delivered on its promise to continuously increasethe Old-Age Pension as well as the Social Package5 for persons with severedisabilities. As of December 2016, the level of the Old-Age Pension benefits wasGEL 180 a month (GEL 216 a month for persons living in high mountainousareas), which was higher than the subsistence minimum level of GEL 162 amonth paid for working-age males. However, the Social Package for survivorsand persons with moderate disabilities has been maintained at the same level(GEL 100 a month), thus its purchasing power has decreased over time due to

5. In 2012, disability and survivor pensions were renamed as the “Social Package”.




58

inflation. Furthermore, because of the considerable depreciation of the nationalcurrency throughout this period, the dollar value of Old-Age Pension and SocialPackage for persons with severe disabilities decreased from USD 73 to USD 68.6

To facilitate the long-term pension reform, the government set up a newdepartment – Capital Market Development and Pension Reform Department –within the Ministry of Economy and Sustainable Development. After nearly fouryears of work, the department produced a blueprint for reform (Ministry ofEconomy and Sustainable Development, 2016). The document identifies theprimary objectives of the reform as being to ensure adequate replacement ratesunder the Old-Age Pension for future pensioners and to ensure fiscalsustainability in the face of an ageing and declining population. To support theseobjectives, in the second half of 2017 the government will introduce a privatepension system, open to all formally-employed individuals. Significantly, incontrast to the pledge made in the election manifesto, membership will bevoluntary and not mandatory. In order to participate, employees will be requiredto contribute 2 per cent of their salaries, which will be matched by equalamounts from both employers and the state. As a defined contribution scheme,the accrued savings will be placed in various investments by a special newly-created non-profit fund, reporting to the Parliament of Georgia. Contributions,as well as returns on investments, will be tax free. Upon retirement, the scheme’sparticipants will be able to withdraw part of their savings as a lump sum andpurchase annuities with the remainder. At the same time, the government willmaintain the existing Old-Age Pension (to be renamed as the “Social Pension”,which will be indexed annually in line with inflation) for those individuals whoseaccrued savings in the private scheme fall below a predefined threshold or whohave no formal employment record.

As in the case of pensions, the coalition government raised the monthly amountof means-tested social assistance (TSA) in 2013 (from GEL 30 for the first memberof the household and GEL 24 for each additional member, to GEL 60 for the firstmember and GEL 48 for each additional member), but has since tried to containfuture expenditures. In 2015, it revised the proxy means-testing formula infavour of families with children and tapered the level of benefits based on theproxy means-testing score.7 It also introduced two additional transfers forchildren. The first is a small top up (GEL 10 a month) that families with a TSAscore below 100,001 receive for each child. The second, the Demographic

6. Author’s own calculations using the official currency exchange rates provided by the National Bankof Georgia (2016).7. Families with a proxy TSA score below 30,001 receive GEL 60 a month per family member; with ascore from 30,001 to 57,000 – GEL 50 a month per family member; with a score from 57,001 to 60,000 –GEL 40 a month per family member; with a score from 60,001 to 65,001 – GEL 30 a month per familymember (SSA, 2016a).




59

Benefit, is to encourage families in regions with negative population growth rates tohave more children by providing GEL 150 to GEL 200 a month (the higher rate ispayable in high mountainous areas) for the third and each subsequent child untilthey reach age two.

With the assistance of the International Labour Office, the coalition majority inthe Parliament also legislated amendments to the labour code in 2013. Some ofthese amendments had long been requested by the global custodian of labourrights.8 The cap on working hours has been reduced from 41 to 40 hours perweek and employers now must pay a higher wage for overtime work. Themaximum working hours for children aged 16–18 and aged 14–16 have been setat 36 hours and 24 hours per week, respectively, and their consent is required forovertime work. Discrimination is prohibited not only at work, but during therecruitment and selection process. Unused leave must be reimbursed and thoseemployed in dangerous occupations are entitled to ten additional days of paidleave. The paid sickness/temporary disability period has been extended from 30 to40 days for a single leave period, and from 50 to 60 days across a calendar year.The maternity leave period has been extended from 477 days to 730 days and thepaid maternity leave period from 126 to 183 days. Furthermore, employers canno longer dismiss employees without a justified reason listed in the code.Employees can demand a justification of dismissal and contest the decision incourt, in which case employers have to carry the burden of proof. The noticeperiod for large-scale redundancies (laying off 100 persons or more) has increasedfrom 30 to 45 days.

The amendments also strengthen the position of the trade unions – a separatechapter on freedom of association has been added to the code. The codeexplicitly prohibits discrimination on the basis of trade union membership andreduces by half the minimum number of employees required to form a tradeunion, from 100 to 50. Also, the 90-day limit on the duration of strikes has beenremoved while the same limit on the duration of lockouts has been maintained.The revised labour code also establishes a Tripartite Social DialogueCommission, headed by the Prime Minister and with government, trade unionsand employers’ associations each represented by six members. Further, in 2015,the Labour Conditions Inspection Department was re-established within theMOLHSA.

These amendments were welcomed by the ILO Committee of Experts on theApplication of Conventions and Recommendations (ILC, 2015) as they certainlyreverse some of the major contractions of employee rights seen under the UNMgovernment and, in theory, should strengthen employees’ bargaining power.However, valid concerns remain about their actual implementation. For example,

8. See for example ILC (2007, 2008, 2010, 2011, 2012, 2013).




60

the real purpose of the Labour Conditions Inspection Department remains elusive,at least for the time being, as an employer’s permission is required for it to enter awork premise to undertake inspection and its recommendations are not legallybinding (EC, 2016). Similarly, the Tripartite Commission is also an advisorybody that aims to facilitate dialogue between social partners and producerecommendations rather than make binding decisions. Moreover, despite itsinauguration in 2013, it convened for the first time in 2016. It adoptedits programme of work at its first session that year, but the two next sessionswere postponed due to the upcoming elections.

The present Georgian social protection system

Social transfers

Categorical and means-tested cash transfers are the main instruments of theGeorgian social protection system. All individuals reaching retirement age (age 60for women; age 65 for men) are entitled to the Old-Age Pension. The entitlementis based solely on age, without any contribution or means-testing requirementand recipients employed in the private sector can continue working without lossof entitlement. The level of transfer is flat rate (GEL 180 a month in 2016), andpensioners living in high mountainous locations receive a supplement of GEL36 a month to compensate for the additional needs arising from residing in aharsh geographical terrain. Instead of the Old-Age Pension, retired senior-levelpublic officials and former personnel of law enforcement agencies receive“Compensations”, which are calculated based on the number of years in publicservice and final salary, but capped at GEL 540 a month.

In addition to the Old-Age Pension, there are several categorical benefits.Disabled persons, orphan children and parents of military personnel killed inaction are entitled to the Social Package, previously called Disability and SurvivorPensions. The level of this transfer varies from GEL 100 to GEL 524 a month,depending on the severity of the assessed disability and whether or not theperson’s breadwinner had died on military duty. IDPs from Abkhazia and SouthOssetia regions can claim the IDP Allowance at GEL 45 a month. As discussedabove, in regions with a declining population, children younger than age 2 areentitled to the Demographic Benefit.

These transfers are not contributory and can be considered as social assistance.However, officially they are classified as “obligatory state transfers”, which isdistinct from “state social assistance” benefits (Parliament of Georgia, 2006). Theprimary social assistance benefit is Targeted Social Assistance (TSA). Familiesthat apply to receive this transfer have to undergo a proxy means-testing




61

procedure. This implies generating the family’s welfare score based on its assets anda number of characteristics (for example, whether or not the family has a membercapable of work, or whether there are disabled persons in the family, etc.). Familieswith a welfare score below 65,001 receive cash assistance, the level of which variesfrom GEL 30 to GEL 60 a month per family member, depending on the score. Inaddition, each family with a score below 100,001 receives GEL 10 a month perchild (see also footnote 7). Another form of social assistance is Utilities Subsidy –the level of this transfer varies from GEL 7 to GEL 44 a month and is intendedfor war veterans, victims of Soviet repressions, persons who have participated inthe emergency response to the Chernobyl (nuclear) catastrophe and several othercategories.

Families that previously had placed a child in state care, and who opt toreintegrate their child back into the family household, receive the ReintegrationAllowance for 12 months, at the rate of GEL 90 a month per reintegrated child.Foster care payments are also classified as social assistance, though these areclearly fees for a service rather than assistance. The foster care payment rangesfrom GEL 450 to GEL 600 a month, depending on whether or not the fosteredchild has an assessed disability.

One distinguishing feature of the Georgian social protection system is theabsence of any form of unemployment benefits. The only employment-relatedstate transfer is maternity benefit, capped at GEL 1,000 for the entire 6-monthperiod of paid maternity leave. Sickness/temporary disability benefits are to bepaid by employers (equal to the average salary in the last three months) as isworkplace injury compensation, if the employer’s responsibility is proven incourt; the court also defines the amount of compensation. Fiscal benefits includeincome tax allowances for disabled people (the first GEL 6,000 of income peryear is not subject to the flat-rate 20 per cent income tax) and families living inhigh mountainous areas (depending on the number of children and employmentsector, from GEL 1,500 to GEL 6,000 of income a year is either fully or partiallyexempted from income tax).

All the benefits outlined above (including Old-Age Pensions) are funded andprovided by the central government. Parallel to these, most municipalities alsoprovide some transfers and in-kind benefits. However, these tend to be limited incoverage and social spending is only a small fraction (5 per cent) of totalmunicipal expenditure (Huby, Bradshaw and Gugushvili, 2010). Moreover, mostof these benefits are one-off and usually targeted at specific categories (disabled,war veterans), though some municipalities use the proxy means-testing score toidentify the beneficiaries (see for example, Tbilisi City Hall, 2016).

The coverage of social transfers can be wide. UNICEF (Mshvidobadze, 2016)estimates that 63 per cent of the population live in households that receive someform of social transfer and 10 per cent receive more than one transfer. Looking at the




62

coverage of individual transfers, the Old-Age Pension has the largest coverage (720,194individuals, corresponding to 19 per cent of the total population),9 followed by theTSA (484,135 beneficiaries) and the Social Package (167,144 beneficiaries) (Table 1).Other benefits have a much lower coverage. For example, only 1,169 womenbenefited from Maternity Pay and 7,224 children received the Demographic Benefit.

To measure the adequacy of the existing cash transfers, different reference pointscan be used, such as GDP per capita (GEL 712 a month in 2015), the average wage(GEL 900 a month in 2015) or the subsistence minimum (GEL 162 a month for aworking age adult in December 2016). Figure 1 presents the average amount ofvarious transfers, with the subsistence minimum offering the most relevantreference point. The average monthly amounts of three transfers – Old-AgePension (GEL 181), Maternity benefit (GEL 164) and Demographic Benefit(GEL 155) – are equal to or slightly higher than the subsistence minimum. Theaverage monthly amounts of TSA (GEL 46), IDP Allowance (GEL 45) and

9. As of 1 January 2016, the population of Georgia is 3,720,400 persons.

Table 1. Social cash transfers

Benefit Eligibility Amount (per month) Coverage

Old-Age Pension Reaching retirement age – age 60 forwomen; age 65 for men

GEL 180 (GEL 216 forpensioners in highmountainous regions)

720,194

Compensations Senior public servants and formerpersonnel of law enforcement agencies

Based on number of yearsin service and final salary,maximum amount GEL 540

21,092

Social Package(previously Disabilityand Survivor Pensions)

Having severe or moderate disability; childhaving lost one of the parents, parents ofmilitary personnel killed in action

GEL 100 to GEL 524 167,144

Utilities Subsidy War veterans, victims of Soviet repressionsand other categories

GEL 7 to GEL 44, dependingon category

28,581

TSA Households with proxy means-testingscores below 65,001

GEL 30 to GEL 60 perfamily member

484,135

IDP Allowance IDPs GEL 45 225,597

Maternity Pay Employed women giving birth or adoptinga child(ren)

GEL 1,000 (maximum forsix months)

1,169

Child ProtectionBenefits (reintegrationand foster care)

Families taking children out of state careservices; foster carers

GEL 90 to GEL 600 1,779

Demographic Benefit Children younger than age 2 living insix regions with declining population

GEL 150 (GEL 200 in highmountainous areas)

7,224

Source: SSA (2016b).




63

Utilities Subsidy (GEL 17) are less than a third of the subsistence minimum, and theSocial Package (GEL 112) is a little more than two-thirds of the subsistenceminimum. The only transfers with an average monthly amount significantlyhigher than the subsistence minimum is the group of Compensations (GEL 416).

Public health and social services

Public health care provision consists of public health insurance and state-fundedhealth programmes. Public health insurance covers either fully or partially acomprehensive set of inpatient and outpatient services, medical tests and childdelivery costs. War veterans and families with a TSA score below 100,000 receivethe most generous insurance packages with minimal co-payment requirements.State-funded programmes include several public health programmes(immunization, early diagnosis and screening, epidemiological monitoring, etc.)and disease-specific programmes (mental health, infectious diseases, tuberculosis,HIV/AIDS, dialysis and kidney transplantation, cardio surgery, etc.).

Social care services are provided to several target groups. These includechildren encountering problems with parental care at home, children and adultswith disabilities, vulnerable mothers with children, older persons, war veterans,

Figure 1. The average amounts of social transfers (GEL a month in October 2016)

Source: SSA (2016b).




64

and victims of domestic violence and trafficking. The services for children whoexperience problems with parental care include foster care, day care centres, smallgroup homes and shelters for homeless children living on the streets. Childrenwith specific disabilities (e.g. severely delayed development, cerebral palsy) arealso entitled to early childhood development and rehabilitation programmes andhome care. To prevent the abandonment of children, families in critical materialconditions may receive one-off support in terms of in-kind benefits (furniture,household equipment, home repairs, etc.), capped at GEL 1,000, and there arealso shelters for vulnerable mothers with children. Disabled adults can benefitfrom community-based services, day care services or residential care. There arealso shelters for older persons and victims of domestic violence and trafficking.War veterans are also entitled to rehabilitation services. The overall coverage ofsocial services is very limited – 5,461 children and adults in total – this is slightlymore than 0.1 per cent of the total population (Table 2).

Table 2. Coverage of social services

Type of care Number of beneficiaries

Foster care 1,355

Daycare centres for non-disabled children 445

Daycare centres for children with disabilities 627

Small group homes 328

Shelters for children living on the street 149

Rehabilitation programme for children with disabilities 428

Early childhood development programme for children with disabilities 523

Shelters for vulnerable mothers with children 53

Shelters for victims of domestic violence and trafficking 213

Emergency support for families in urgent material hardship 245

Home care service for children with severely delayed development 39

Daycare centres for adults with disabilities 456

Community-based services for adults with disabilities 204

Residential care for adults with disabilities 189

Residential care for older persons 194

Rehabilitation of war veterans 13

Total 5,461

Source: SSA (2016b) and State Fund for Protection and Assistance of Victims of Human Trafficking (2016).




65

Labour legislation

Georgia is a state party to each of the International LabourOrganization’s (ILO) eightfundamental Conventions. As noted, labour relations are regulated by the labourcode, last revised in 2013. The code defines the minimum working age (16 years),the working-hours limit (40 hours a week), overtime pay, paid and unpaid annualleave (at least 24 days a year, and at least 15 days a year, respectively), limitations onhazardous work for vulnerable employees (women and children younger thanage 18), maternity leave (730 days, of which 183 are paid) and maternity pay,occupational health and safety provisions, qualified reasons for dismissal, dismissalnotice period (30 days in case of an individual dismissal, 45 days in case of a massredundancy) and severance pay (one month’s salary if the notice period is 30 days,or two months’ salary if the notice period is less than 30 days).

The labour code does not currently include a number of important guaranteesto employees, such as a minimum wage and specific clauses prohibitingdiscrimination against women. The main vehicle for enforcing the code is thecivil court as there is no specialized labour court or tribunal. Further revisions ofthe labour code are on the agenda of the Tripartite Social Dialogue Commission,and informal tripartite discussions started with ILO assistance in June 2016.

Financing and administration of welfare

In 2016, total social spending in Georgia on health and social protection (excludingeducation) was GEL 3.2 billion, corresponding to 32 per cent of the state budgetand 9.4 per cent of GDP (Figure 2).

The overall funding and administration of social programmes remains highlycentralized. The main directions of economic and social policy are defined bythe Cabinet of Ministers, which is led by the Prime Minister and reports to theParliament.10 The Parliament also makes amendments to, and adopts, the statebudget. The Ministry of Finance (MOF) is responsible for raising revenues anddefining the overall spending ceilings, which are then negotiated with the lineministries. The MOLHSA administers social transfers and in-kind benefits, socialcare services and health programmes through its subsidiary bodies, including theSocial Services Agency (SSA) which has district-level branches. The Ministry ofRefugees and Accommodation (MORAA) deals with IDP issues, including theprovision of housing expenses. Local authorities design and fund municipalsocial programmes, including pre-school services, but they also get some fundingfrom the MOF through equalizing transfers.

10. Before the constitutional amendments came into force in 2012, the Prime Minister wassubordinated to the President.




66

Comparison with other welfare states

As a part of the Soviet Union prior to the transition, Georgia had a highlydistinctive communist-type welfare regime. To replace this with a welfare regimesuited to high- and middle-income countries continues to present Georgia withmajor challenges. The context and task is completely different from that associatedwith “informal-conservative” Latin American (Barrientos, 2004) and “productive”East Asian welfare (Kwon, 1997) regimes. There are also very few, if any,similarities with the Mediterranean welfare model (Ferrera, 1996). To some extent,the Georgian regime does resemble the liberal welfare regime (see for exampleEsping-Andersen, 1990), with a low level of spending, considerable reliance onmeans-testing, significant share of private spending on health care, low level ofde-commodification, and market-based stratification. However, if placed withinthis cluster, the Georgian welfare regime would be an extreme outlier since eventhe least generous liberal welfare regimes (United States, Australia, UnitedKingdom) direct a much larger share of GDP to social expenditures and havemuch more comprehensive welfare arrangements. For example, unlike thesewelfare regimes, Georgia has no unemployment transfers, active labour market

Figure 2. Health and social protection spending in 2016 (million GEL)

Source: Ministry of Finance (2016).




67

policies, or housing benefits. In addition, social care services have a very lowcoverage rate, at around 0.1 per cent of total population.

There are very few comparative quantitative studies that locate the Georgianwelfare regime within the broader context of European welfare states. This maybe due to the shortage of reliable data. A first study was carried out by Fenger(2007), who deployed hierarchical cluster analysis to identify distinctive groupsof welfare regimes among former communist countries. Georgia, together withMoldova and Romania, formed a separate cluster that was clearly lagging behindother clusters in terms of both social spending and general socio-economicindicators. Hence, Fenger used the term “developing welfare state type” to definethis regime type, suggesting that it is still evolving towards the regime-typeassociated with “mature welfare states” (Fenger, 2007, p. 25).

Another study was conducted by Gugushvili and Sundberg (2012) who soughtto measure the impact of economic globalization on welfare systems in theformer communist countries of Central and Eastern Europe and the SovietUnion. Using hierarchical cluster analysis, they clustered Georgia in the “East”group of countries together with Armenia, Azerbaijan, Kazakhstan andKyrgyzstan. Compared to the other three identified geographical clusters(Northern, Southern, Western), this group was distinguished by a low level ofsocial spending and income taxes, flexible labour market regulations, a low levelof de-commodification, high inequality, and poor health and education outcomes.

Strengths and weaknesses of the present Georgian socialprotection system

Despite these comparative assessments, it should be emphasized that the Georgiansocial protection system has a number of strengths. These include the broadcoverage of social transfers, simple eligibility criteria, and a well-functioningadministrative infrastructure. The eligibility criteria for all benefits, with theexception of TSA, is clear, and the application process is simple and fast. Allsocial transfers are funded from a central budget, administered by a single agency– the SSA – and promptly delivered through the banking system. Since 2003there have been no arrears or any delays for any of the transfers and the overalladministration costs are low. Increasing reliance on means-testing and theabsence of contributory benefits improves redistribution and helps to channelmore resources to those in need (Gugushvili, 2014).

Yet, the system suffers from a number of significant deficiencies that limit itseffectiveness and efficiency. The main drawback is the low level of spending onall its main components. As a share of GDP, Georgia spends 46 per cent of theaverage OECD country expenditure for social protection and 37 per cent of average




68

OECD country expenditure on health. Overall, among OECD countries, onlyMexico has a lower level of total expenditure (Figure 3). Furthermore, the absenceof some core benefits (such as unemployment benefits and housing transfers) andthe low level of transfers render the system unable to offer the majority of thepopulation protection from main social risks, such as unemployment, sicknessand old age. Owing to ad-hoc indexation, the real purchasing power of benefitssignificantly fluctuates across different years. Redistribution effects are also limitedby the universality of the main transfer, the Old-Age Pension, since takeup rates arealmost universal across all quintiles (Gugushvili, 2014). Similarly, a heavy relianceon categorical targeting does not improve redistribution significantly, especiallywhen the target groups are selected on the basis of particular merits rather thantheir level of vulnerability.

The TSA is the most redistributive among the existing transfers, however is notwithout problems. Given that only a small share of Georgian families rely solely onpaid employment, it is difficult to capture households’ real incomes orconsumption. For this reason the SSA uses families’ assets (mostly durable items)as the main proxies for measuring welfare, but as practice shows these are rather

Figure 3. Health and social protection spending as a share of GDP in Georgia andOECD countries

Note: The data for Georgia is from 2016, while for OECD countries the latest comparable data is from 2014.

Sources: Geostat (2016) and OECD (2016).




69

inaccurate indicators. The proxy means-testing formula itself is very complicatedand the majority of applicants do not understand how their scores are derived.The lines beyond which different benefits are granted are absolutely arbitrary;they are not related to any (relatively) objective criteria. Rather, it seems thatthese thresholds are defined by the amount of available funding. As aconsequence of this narrow targeting, a large proportion of poor families aredisqualified, which may cause their legitimate frustration with the system.

The low levels of most transfers and limited coverage of poor working-agepersons also compromises the system’s effect on poverty. Recent studies showthat social transfers in Georgia reduce absolute poverty by between a quarter anda third, depending on the poverty thresholds used (Mshvidobadze, 2016;Gugushvili, 2014). This is not particularly impressive, especially given that theabsolute poverty thresholds used in these studies are rather low.

Despite the recent changes to the labour code, the protection of employees’ rightsstill remains a major challenge. The Labour Conditions Inspection Department’sfunctions are still at a nascent stage of development. The number of occupationalinjuries and deaths remains very high – 124 serious injuries in 2015, of which42 were fatal (Public Defender of Georgia, 2015). The national ombudsman’s office(Public Defender of Georgia, 2014, 2015) found multiple cases of unjustifieddismissal and other violations of employee rights, including in the public sector.Moreover, the Georgian Trade Union Confederation reported numerous cases ofdiscrimination against trade union members in its report to the ILO Committeeof Experts on the Application of Conventions and Recommendations (ILC, 2015).

Discussion:What next for the Georgian social protection system?

Despite the major contraction experienced in the first decade of transition,the Georgian social protection system is now in a phase of expansion, and the2013–2016 reforms, especially to universalize public health insurance, arewelcome steps towards its transformation to a modern welfare system. Yet majorchallenges remain, especially in relation to the system’s limited effects onwidespread absolute poverty. To address these challenges will require a moresubstantive expansion of the social protection system as regards the developmentof new programmes and mechanisms and higher spending, as well asimplementing revised labour regulations.

Nevertheless, it is unclear whether the current expansion phase will continuein the immediate future or what shape the future welfare system will take. In thelong run, it may seem reasonable to expect that the Georgian social protectionsystem will follow the same trajectory as that of European welfare systems in thepast – expanding linearly for many years in tandem with economic development,




70

before undergoing significant recalibration in the face of declining/stagnant growthand fundamental socio-economic changes such as demographic changes,globalization, and de-industrialization and the rise in importance of the servicesector, etc. (Taylor-Gooby, 2002). However, in the short to medium term thiscannot be taken for granted for a number of reasons. The main obstacle is thatGeorgia has nothing even resembling the strong trade unions and powerfulleft-wing parties that played a crucial role in developing and, later, defending theextensive welfare systems in Western countries (Georgia’s situation is similar tomany other post-Soviet countries). Georgian trade unions have limitedpolitical power and political parties lack clear ideological stances. It is rarefor political actors to have a consistent socio-economic development strategy oragenda. One exception is the neoliberal UNM, though since moving intoopposition it appears to have relaxed its radical neoliberal views.

The Georgian Dream Party, which won a large victory in the 2016parliamentary elections (with 115 out of 150 parliamentary seats, aconstitutional majority), is no exception to this general trend. While identifyingitself as social democratic and having implemented a number of progressivesocial protection measures, the Georgian Dream government’s (Government ofGeorgia, 2016) new four-point action plan is strongly market-oriented. It pledgesnot to increase the “state’s pressure on the economy” (p. 12; author’s translation)and promises to “liberalize the business environment” (p. 15; author’stranslation) through announcing a moratorium on establishing new stateregulators, providing tax incentives for investors, extending free trade agreements(notably with the People’s Republic of China), and relaxing fines for minoreconomic fraud and tax evasion. Commitments in the social protection sphereare few and vague, such as refining the targeting of social assistance, developingnew benefits for families with many children, and improving coordinationbetween employers and jobseekers. More specific commitments include theimplementation of pension reform and the purchasing of housing for recipientsof IDPs and environmental migrants.

Another considerable risk is the significant deceleration of economic growthsince 2012. In the past four years growth has averaged 3.4 per cent, almost halfof the 6.2 per cent achieved across 2004–2012. Although the rate of growth hashad little direct effect on poverty (Gugushvili, 2017), economic growthcombined with strong anti-corruption measures has facilitated an exponentialincrease in public revenues – a considerable portion of which was spent on thereanimation of the social protection system. The slowdown in growth isoccurring in the context of a rapidly ageing society, however. Among transitioncountries, Georgia has one of the worst dependency ratios, with the ratioprojected to worsen in the next decade (World Bank, 2007). This will drive upthe costs of two major social protection programmes – Old-Age Pensions (even




71

after the implementation of the planned reform) and public health insurance –

which account for around three-fifths of social spending. And without a firmcommitment to allocate an increasing share of national income to socialspending, the risk is that even the existing minimal commitments will begradually watered down. In point of fact, in 2016 the government experiencedsignificant problems with meeting the costs of those programmes (Institute forDevelopment of Freedom of Information, 2016; Radio Liberty, 2016), and sofar its response has not been encouraging. In relation to Old-Age Pensions, itchose to cut by stealth – allowing a major currency devaluation to erode thepurchasing power of the benefit despite a small indexed increase. In relation touniversal health insurance, its approach is more problematic – a long waitingperiod has been established for privately-insured citizens before they can shiftto public insurance, thereby generating a classic example of an adverse incentivethat will undoubtedly erode public trust in the public scheme (and thereforethe willingness to pay for it though taxes).

A further potential handicap is the unlikelihood of Georgia joining theEuropean Union (EU) in the foreseeable future. The prospect of EU membershipwas a factor that played a significant role in the development of welfare systemsin Central and Eastern European countries that eventually did become EU MemberStates. As one of the key conditionalities for joining the EU, Orenstein and Haas(2005) point out that parties on both sides of the political spectrum in aspirantcountries quickly grasped that they had to ensure welfare spending and provisioncomparable to Western European countries. This was necessary specifically tolimit the risks of social dumping and mass migration that other Member Statesfeared. Strong technical support from the EU, for example through the twinningof existing and prospective Member States’ social welfare ministries, furthercontributed to the development of these welfare systems by allowing them to tapinto top level expertise and experience.

At present, there are no such pull factors for Georgia. This significantly reducesGeorgian politicians’ appetites for embarking on a complex reform agenda that willinevitable require a considerable increase in public spending (and thus in taxation),which in turn is likely to continue to cause reservations among and triggerresistance from business elites. This feeds a degree of reluctance on the part ofthe Georgian government to engage with any arguments or lobbying presentedby the EU in favour of the substantive reform of the welfare system. That said,resistance is currently not as strong as it was during the UNM administration’speriod in office. The EU and the ILO have played a considerable role in promotingsocial reforms and the revision of the labour code.

The risks discussed here are significant, but there also reasons for optimism.Once enacted and implemented, welfare entitlements are very difficult to reversein democratic states. The major retrenchment of the Georgian social protection




72

system in the early 1990s occurred against the background of the near collapse ofthe state. The Georgian electorate came to understand that it was impossible tomaintain Soviet-style entitlements in the context of a practically bankrupt statethat was still recovering from the impacts of major armed conflicts and theunprecedented transformation of an already stagnant economy. Generally for allcountries, in the absence of such dramatic events, to attempt to cut backsignificantly on any existing national programmes might be expected to exposeany ruling political party to a major backlash from voters.

There is also a welcome change in public attitudes as regards the social contractbetween the state and the people. Unlike the generally passive Soviet citizens, whowere often characterized as having being grateful for whatever benefits the“benevolent” government would pass down, modern Georgian citizens increasinglyperceive their relation with the state in contractual terms. They demand tangiblereturns on the taxes they pay, whether this be modern infrastructure or adequateprotection from social risks at different stages of the life course.

Finally, there is a slow but growing recognition among Georgian policy-makersthat growth alone is not sufficient to tackle the huge social problems that thecountry faces. The Georgian political mainstream, as well as senior publicservants, once simplistically believed that rapid growth would automaticallybenefit everyone – even if some were to benefit more than others. Georgia’s poorrecord in poverty reduction all too obviously debunked this fallacy. Having nowlearnt the hard way that a redistributive social protection system is critical forinclusive growth, this should make future Georgian governments moresympathetic to the idea of the modern welfare state. Therefore, the furtherexpansion of the Georgian welfare system may yet become a more likely possibility.

Bibliography

Aidukaite, J. 2004. The emergence of the post-socialist welfare state: The case of the Baltic

States: Estonia, Latvia and Lithuania. Stockholm, Södertörns Högskola.

Atkinson, A. B.; Micklewright, J. 1992. Economic transformation in Eastern Europe and the

distribution of income. Cambridge, Cambridge University Press.

Barrientos, A. 2004. “Latin America: Towards a liberal-informal welfare regime”, in

I. Gough and G. D. Wood (eds), Insecurity and welfare regimes in Asia, Africa and

Latin America: Social policy in development context. Cambridge, Cambridge University Press.

Chandler, A. 2001. “Globalization, social welfare reform and democratic identity in Russia

and other post-communist countries”, in Global Social Policy, Vol. 1, No. 3.

Cook, L. J. 2010. “Eastern Europe and Russia”, in F. G. Castles (ed.), The Oxford handbook

of the welfare state. Oxford, Oxford University Press.




73

https://www.diva-portal.org/smash/get/diva2:16651/FULLTEXT01.pdf

https://www.diva-portal.org/smash/get/diva2:16651/FULLTEXT01.pdf

Democratic Republic of Georgia. 1921. Constitution of Georgia, 1921. Tbilisi.

Deacon, B. 1992. “East European welfare: Past, present and future in comparative context”,

in B. Deacon (ed.), The new Eastern Europe: Social policy past, present and future. London,

Sage.

Deacon, B. 1993. “Developments in Eastern European social policy”, in C. Jones (ed.), New

perspectives on the welfare state in Europe. London, Routledge.

Deacon, B. 2000. “Eastern European welfare states: The impact of the politics of globali-

zation”, in Journal of European Social Policy, Vol. 10, No. 2.

EC. 2016. The EU special incentive arrangement for sustainable development and good

governance (‘GSP+’) covering the period 2014–2015 (Joint Staff working document).

Brussels, European Commission.

Esping-Andersen, G. 1990. The three worlds of welfare capitalism. Cambridge, Polity.

Fajth, G. 1999. “Social security in a rapidly changing environment: The case of the post-

communist transformation”, in Social Policy and Administration, Vol. 33, No. 4.

Falkingham, J. 2003. Inequality and poverty in CIS-7 1989-2002 (SSRC Applications and

Policy working paper, No. A03/02). Southampton, University of Southampton – Social

Statistics Research Centre.

Fenger, H. J. M. 2007. “Welfare regimes in Central and Eastern Europe: Incorporating post-

communist countries in a welfare regime typology”, in Contemporary Issues and Ideas in

Social Sciences, Vol. 3, No. 2.

Ferrera, M. 1996. “The ‘southern model’ of welfare in social Europe”, in Journal of

European Social Policy, Vol. 6, No. 1.

Gamkrelidze, A. et al. 2002. Health care systems in transition: Georgia (European

Observatory on Health Care Systems, Vol. 4, No. 2). Copenhagen, World Health

Organization Regional Office for Europe – European Observatory on Health Care Systems.

Georgian Dream. 2012. Election programme. Tbilisi.

Geostat. 2016. Government finance statistics. Tbilisi, National Statistics Office of

Georgia.

Government of Georgia. 2007. ქვეყნის ძირითადი მონაცემები და მიმართულებები

2008-2011 წლებისათვის [Basic data and directions 2008–2011]. Tbilisi, Ministry of

Finance.

Government of Georgia. 2016. სამთავრობო პროგრამა 2016-2020: თავისუფლება,

სწრაფიგანვითარება, კეთილდღეობა [Government Plan 2016–2020: Liberty, rapid

development, welfare]. Tbilisi.

Gugushvili, D. 2014. Do the benefits of growth trickle-down to Georgia’s poor? A case for a

strong welfare system (PhD Thesis). Canterbury, University of Kent.

Gugushvili, D. 2017. “Lessons from Georgia’s neoliberal experiment: A rising tide does not

lift all boats”, in Communist and Post-Communist Studies, Vol. 50, No. 1.

Gugushvili, D.; Sundberg, T. 2012. The post-Soviet welfare state: Racing to the bottom? A case

for divergence (conference paper, Social Policy Association Conference, University of

York, 16–18 July). London.




74

https://matiane.wordpress.com/2012/09/04/constitution-of-georgia-1921/

https://eeas.europa.eu/sites/eeas/files/european_commission._2016._report_on_the_generalised_scheme_of_preferences_during_the_period_2014-2015.pdf

https://eeas.europa.eu/sites/eeas/files/european_commission._2016._report_on_the_generalised_scheme_of_preferences_during_the_period_2014-2015.pdf

https://pdfs.semanticscholar.org/d202/395f3c05b32ce29c3c159540f8734f012cc6.pdf

http://www.euro.who.int/__data/assets/pdf_file/0017/80702/e75489.pdf

http://www.gd.ge

http://geostat.ge/index.php?action=page&p_id=314&lang=eng

http://old.mof.ge/BDD

http://old.mof.ge/BDD

http://gov.ge/index.php?lang_id=GEO&sec_id=68&info_id=58446

http://www.sciencedirect.com/science/article/pii/S0967067X16300538

http://www.sciencedirect.com/science/article/pii/S0967067X16300538

Huby, M. et al. 2012. The well-being of children and their families in Georgia: Georgia Welfare

Monitoring Survey Second Stage 2011. Tbilisi, UNICEFGeorgia; Heslington, University of York.

Huby, M.; Bradshaw, J.; Gugushvili, D. 2010. How do Georgian children and their families

cope with the impact of the financial crisis? Report on the Georgia Welfare Monitoring

Survey, 2009. Tbilisi, UNICEF Georgia; Heslington, University of York.

ILC. 2007. Report of the Committee of Experts on the Application of Conventions and

Recommendations (Report III (Part 1A), International Labour Conference,

96th session). Geneva, International Labour Office.











Recommendations (Report III (Part 1A), International Labour Conference, 101st session).

Geneva, International Labour Office.


Recommendations (Report III (Part 1A) International Labour Conference, 102nd

session). Geneva, International Labour Office.


Recommendations (Report III (Part 1A) International Labour Conference,


Institute for Development of Freedom of Information. 2016. საყოველთაო ჯანდაცვის

პროგრამის გამოწვევები და მათი დაძლევის გზები [Challenges of the universal

health insurance programme and ways to solve it]. Tbilisi.

Kwon, H. 1997. “Beyond European welfare regimes: Comparative perspectives on East

Asian welfare systems”, in Journal of Social Policy, Vol. 26, No. 4.

International Trade Union Confederation. 2008. “Georgia: Labour code tears

fundamental rights to shreds”, in Union View, No. 9.

Manning, N. 1992. “Social policy in the Soviet Union and its successors”, in B. Deacon

(ed.), The new eastern Europe: Social policy past, present and future. London, Sage.

Milanovic, B. 1995. Poverty, inequality, and social policy in transition economies (Policy

research working paper, No. 1530). Washington, DC, World Bank.

Ministry of Economy and Sustainable Development. 2016. საქართველოს საპენსიო

რეფორმა: არსებული საპენსიო სისტემის მოდიფიკაცია და კერძო დაგროვებითი

საპენსიო სისტემის დანერგვა [Pension reform in Georgia: Reforming the existing

pension system and establishing a private pension system]. Tbilisi.




75

http://www.unicef.org/socialpolicy/files/Georgian_Families_and_Their_Children_Cope_w_Financial_Crisis.pdf



http://www.ilo.org/public/libdoc/ilo/P/09661/09661(2007)1A.pdf

http://www.ilo.org/public/libdoc/ilo/P/09661/09661(2007)1A.pdf

http://www.ilo.org/public/libdoc/ilo/P/09661/09661(2008-97-1A).pdf












https://idfi.ge/public/upload/Meri/Policy%20Brief.pdf

https://idfi.ge/public/upload/Meri/Policy%20Brief.pdf

http://www.ituc-csi.org/georgia-labour-code-tears

http://www.ituc-csi.org/georgia-labour-code-tears

http://elibrary.worldbank.org/doi/abs/10.1596/1813-9450-1530

http://www.economy.ge/uploads/meniu_publikaciebi/ouer/Goverment_Report_on_pensions_Final_GEO.pdf



Ministry of Finance. 2016. State budget 2016. Tbilisi.

Muller, A. 2012. Employment protection legislation (EPL) of Georgia: A review based on ILO

standards, OECD indicators and comparative labour law (EPLex In Brief, No. 1). Geneva,

International Labour Office.

Mshvidobadze, A. 2016. The well-being of children and their families in Georgia. Tbilisi,

UNICEF Georgia.

National Bank of Georgia. 2016. Statistical data. Tbilisi.

O’Brien, C.; Chanturidze, T. 2009. UNICEF: Assessment of the child welfare reform process in

Georgia. Tbilisi, UNICEF Georgia.

Orenstein, M. A. 2008. “Post-communist welfare states”, in Journal of Democracy, Vol. 19,

No.4.

Orenstein, M. A.; Haas, M. R. 2005. “Globalization and the future of welfare states in post-

communist east-central European countries”, in M. Glatzer and D. Rueschemeyer (eds),

Globalization and the future of the welfare state. Pittsburgh, PA, Pittsburgh University

Press.

OECD. 2016. Social expenditure database SOCX. Paris, Organisation for Economic Co-

operation and Development.

OSGF. 2012. Independent Georgia: Health and social protection systems – Analytical review.

Tbilisi, Open Society Georgia Foundation.

Parliament of Georgia. 2006. კანონი სახელმწიფო სოციალური დახმარების შესახებ

[Law on social assistance]. Tbilisi.

Public Defender of Georgia. 2014. Annual report of the Public Defender of Georgia: The

situation of human rights and freedoms in Georgia. Tbilisi.

Public Defender of Georgia. 2015. Annual report of the Public Defender of Georgia: The

situation of human rights and freedoms in Georgia. Tbilisi.

Radio Liberty. 2016. “საყოველთაოჯანდაცვის” საყოველთაოპრობლემები [Universal

problems of “universal health insurance”]. Prague.

Shvelidze, Z. 2012. Transition from Soviet to liberal labour law: Labour standards in Georgia.

Tbilisi, Ivane Javakhishvili Tbilisi State University – Law Faculty.

SSA. 2016a. Targeted social assistance. Tbilisi, Social Services Agency.

SSA. 2016b. Statistical data. Tbilisi, Social Services Agency.

Standing, G. 1996. “Social protection in Central and Eastern Europe: A tale of slipping

anchors and torn safety nets”, in G. Esping-Andersen (ed.) Welfare states in transition:

National adaptations in global economies, London, Sage.

State Fund for Protection and Assistance of Victims of Human Trafficking. 2016.

სტატისტიკა [Statistics]. Tbilisi.

Taylor-Gooby, P. 2002. “The silver age of the welfare state: perspectives on resilience”, in

Journal of Social Policy, Vol. 31, No. 4.

Tbilisi City Hall. 2016. Social programmes. Tbilisi.

Tvalchrelidze, A. 2003. “The social security system in Georgia and possible trends in its

development”, in IDEA, Democratization in Georgia: Economic transformation and social




76

http://mof.ge/en/4890

http://www.ilo.org/ifpdial/information-resources/publications/WCMS_202301/lang--en/index.htm

http://www.ilo.org/ifpdial/information-resources/publications/WCMS_202301/lang--en/index.htm

https://www.nbg.gov.ge/index.php?m=304&lng=eng

http://www.bettercarenetwork.org/sites/default/files/UNICEF - Assessment of the Child Welfare Reform Process in Georgia.pdf

http://www.bettercarenetwork.org/sites/default/files/UNICEF - Assessment of the Child Welfare Reform Process in Georgia.pdf

http://scholar.google.com/scholar?cluster=3051549064821888243&hl=en&oi=scholarr

http://scholar.google.com/scholar?cluster=3051549064821888243&hl=en&oi=scholarr

http://www.oecd.org/social/expenditure.htm

https://www.osgf.ge/files/2013/publikaciebi 2013/Report_Color_ENG_WEB.pdf

http://ssa.gov.ge/index.php?lang_id=GEO&sec_id=722

http://ombudsman.ge/uploads/other/3/3510.pdf




http://www.radiotavisupleba.ge/a/sakoveltao-jandacvis-sakoveltao-problemebi/27683758.html

http://www.radiotavisupleba.ge/a/sakoveltao-jandacvis-sakoveltao-problemebi/27683758.html

http://ilera2012.wharton.upenn.edu/RefereedPapers/ShvelidzeZakaria ILERA updated.pdf



http://atipfund.gov.ge/index.php

http://tbilisi.gov.ge/?lang=en

http://www.idea.int/publications/georgia/upload/Book-08_scr.pdf

http://www.idea.int/publications/georgia/upload/Book-08_scr.pdf

security (Building democracy in Georgia: Discussion paper, No. 8). Stockholm, Institute

for Democracy and Electoral Assistance.

WHO. 2009. Georgia: Health system performance assessment. Copenhagen, World Health

Organization Regional Office for Europe.

World Bank. 2007. From red to gray: The “third transition” of ageing populations in Eastern

Europe and the former Soviet Union, Washington, DC.

World Bank. 2012. Fighting corruption in public services: Chronicling Georgia’s reforms

(Directions in development). Washington, DC.

World Bank. 2016. Health expenditure, public (% of GDP). Washington, DC.




77

http://www.euro.who.int/__data/assets/pdf_file/0012/43311/E92960.pdf

http://siteresources.worldbank.org/ECAEXT/Resources/publications/454763-1181939083693/full_report.pdf

http://siteresources.worldbank.org/ECAEXT/Resources/publications/454763-1181939083693/full_report.pdf

http://www-wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2012/01/20/000356161_20120120010932/Rendered/PDF/664490PUB0EPI0065774B09780821394755.pdf

http://data.worldbank.org/indicator/SH.XPD.PUBL.ZS

WORLDVIEW

The challenge of pensionreform in Georgia:

Non-contributory pensionsand elderly poverty

Tamila Nutsubidze and Khatuna Nutsubidze

Healthcare and Social Issues Committee,

Parliament of Georgia, Georgia

Abstract Georgia’s national social security system offersalmost universal non-contributory basic pension coverage.The basic pension has, to date, proved effective in dealingwith old-age poverty. But Georgia’s fiscal constraints andageing population also highlight the importance of improvingthe pension system, in order to ensure its sustainability.This article presents policy reform choices, which suggestthat, in Georgia, pension reform might include increasing thestatutory retirement ages and reducing the generosity ofbenefits through means testing. The case of the Georgiannon-contributory basic pension might hold value for somelow- and middle-income countries that are consideringthe implementation of, or expanding coverage under, anon-contributory pension programme.

Keywords social security reform, non-contributory scheme,pension scheme, poverty, Georgia

Addresses for correspondence: Tamila Nutsubidze, Consultant, Healthcare and Social Issues Committee,Parliament of Georgia, 26 Irakli Abashidze St., Kutaisi, Georgia, 4600; email: [email protected] Nutsubidze, Consultant, Healthcare and Social Issues Committee, Parliament of Georgia, 26Irakli Abashidze St., Kutaisi, Georgia, 4600; email: [email protected] authors thank Alicia H. Munnell, the Peter F. Drucker Professor of Management Sciences at BostonCollege’s Carroll School of Management and Director of the Center for Retirement Research atBoston College, Anthony Webb, a Senior Research Economist at the Center for Retirement Researchat Boston College, and Mauricio Soto, a Senior Economist of the Expenditure Policy Division of theInternational Monetary Fund’s Fiscal Affairs Department for comments. They also thank the threeanonymous referees for helpful comments and suggestions.The research reported herein was pursuant to a grant from the Fulbright Visiting Scholar Program,which was funded by the US Department of State’s Bureau of Education and Cultural Affairs.



79

Introduction

As is the case in many low- and middle-income countries, the provision of anadequate old-age income, which includes the objective of mitigating povertyamong the elderly and their households, is a major challenge in Georgia.1 In2006, Georgia introduced a basic non-contributory pension programme,which is financed with tax revenues.2 The impetus for the implementationof the basic pension was the collapse of the Soviet Union in 1991. Dramaticchanges in the country – economic deterioration provoked by the decline inindustrial and agricultural production, a challenging transition to a marketeconomy, military conflicts, rising levels of informal economic activity,migration of the younger population, and population ageing – caused a highincidence of poverty that had to be addressed by social policy. Thenon-contributory basic pension has proved effective in dealing with elderly,household and child poverty in Georgia. However, two dimensions of thenon-contributory pension limit its potential sustainability: the country’s ageingpopulation and scarce public financial resources, both of which impact uponother pressing needs in social policy provision, including for social assistanceand health care.

Since 2012, the Georgian government has put forth a systemic reformproposal – compulsory pension insurance – that includes the implementationof contributory private pension schemes to supplement the basicnon-contributory pension (Government of Georgia, 2012, 2013, 2014a, 2014band 2015; MoF, various years).3 While the proposed reform holds somepromise, it alone would not be sufficient to ensure that the government’s fiscalburden remains reasonable, while also maintaining the basic pension system’s

1. Economic literature suggests that many middle-income countries face the challenge of improvingsustainability in their pension systems, while low-income countries, with younger populations lackadequate coverage for their elderly (Holzmann, Robalino, and Takayama, 2009). Additionally,emerging European countries face the dual challenges of improving the sustainability of pensionsystems while maintaining adequacy (Clements, Eich, and Gupta, 2014; IMF, 2011). Georgia is amongthe lower-middle-income countries with gross national income per capita of 7,040 (PPP, currentinternational dollars, 2013) (World Bank, 2015a).2. This article uses the term “basic pension” derived from the World Bank’s multi-pillar pensiontaxonomy. The World Bank’s pension policy framework flexibly applies a five-pillar model definingthe range of design elements to determine the pension system modalities and reform options thatshould be considered. The five pillars are: a non-contributory “zero pillar”, a mandatory “first pillar”,a mandatory “second pillar”, a voluntary “third pillar”, and a non-financial “fourth pillar”(Holzmann, Hinz, and Dorfman, 2008). This structure is an expansion of the World Bank’s originalconcept of a specific three-pillar pension structure to include two additional pillars: the basic (zero)pillar and non-financial fourth pillar. Additionally, the basic (zero) pillar deals more explicitly withthe poverty objective (Holzmann and Hinz, 2005).3. Reference is also found in Georgian State Budget legislation in 2013, 2014 and 2015.

The challenge of pension reform in Georgia



80

long-term adequacy and creating a system with sufficient flexibility to adapt tothe evolving demographic environment.

This article presents policy reform scenarios that suggest that Georgia’s pensionreform might also include increasing the statutory retirement ages and reducing thegenerosity of benefits through means testing. These reform options would involvedifficult trade-offs, which are also described in this article.

Given the rising international interest in non-contributory pensions– particularly across the last two decades – this article seeks to contribute to theeconomic literature by describing these pensions, their effect in alleviatingpoverty, and the reforms associated with them. Additionally, this article mighthold value for low- and middle-income countries that are considering theimplementation of, or the expansion of coverage under, a non-contributorypension programme. It is worth mentioning that Palacios and Knox-Vydmanov(2014) suggest that most pension schemes in Central Asian and Eastern Europeancountries have a distinct nature that reflects the Soviet history of their pensionsystems. Specifically, they operate pension systems based on payroll contributionsthat have close to universal coverage, and non-contributory pensions are oftenlimited to a tiny minority of the population (for example, in Kyrgyzstan andMoldova, fewer than 2 per cent of population aged 60+ receives a non-contributorypension). Palacios and Knox-Vydmanov (2014) highlight that the proportion of theworkforce in the region that contributes to a pension is decreasing rapidly andsuggest that non-contributory pensions will become increasingly important ifthese countries’ pension systems are to maintain universal coverage.4

The remainder of this article is structured as follows. The next section presents adiscussion of the importance of non-contributory basic pensions in the reduction ofpoverty in Georgia. This includes a brief discussion of the social, economic anddemographic conditions surrounding the formation of the non-contributory basicpension and its role and scope in poverty alleviation. We also review the relevantliterature, with a particular focus on studies that consider the impact of pensionson poverty. The article then examines the current retirement income landscape inGeorgia and identifies fiscal constraints related to the basic pension system.Demographic and basic pension spending projections are then presented. This isfollowed by a discussion of further reform options that could address Georgia’sfiscal and demographic challenges and which are consistent with the need tomaintain the sustainability and adequacy of the non-contributory basic pensionsystem. This leads to the presentation of a number of critical considerations andconclusions about the pension reform options available to Georgia.

4. Palacios and Knox-Vydmanov (2014) point out that Georgia is one country that appears to havealready recognized this trend by having converted its contributory social insurance system into a basicuniversal non-contributory pension in 2006.




81

The methodological approach used in this study involves a detailedunderstanding of the implementation, structure, scope, and administration of thenon-contributory basic pension in Georgia. The research involved collectingand collating country information, qualitative and quantitative data analysis, andinterviews. Projected pension spending was calculated based on the most recentavailable pension spending data (2014) from the Social Service Agency (SSA) ofGeorgia, demographic data from the United Nations and the SSA, and labourforce data from the International Labour Organization (ILO).

More specifically, the basic pension spending projections use a simple modelthat takes into account the evolution of demographics projected by the UnitedNations (ESA, 2015). The model relies on an identity, wherein pension spendingas a per cent of gross domestic product (GDP) is equal to the product of the old-agedependency ratio (the ratio of population aged 60 or older to the populationaged 15–59); the coverage ratio (the ratio of the number of pensioners to thepopulation aged 60 or older); the benefit ratio (the ratio of the average benefit asa per cent of GDP per worker); and the inverse of the employment to populationratio for ages 15–59 (employment is assumed to follow labour force participationand it is obtained from the ILO). In the baseline pension spending projection, it isassumed that the coverage and benefit ratios remain roughly constant (pensionernumbers are projected to grow in line with the old-age population – aged 60 orolder for men and aged 55 or older for women), the employment to populationratio increases slightly, and thus largely reflects the projected evolution of theold-age dependency ratio. In this basic model, changes in the retirement age affectthe coverage ratio – it is assumed that a one-year increase in the retirement agereduces the number of pensioners by the share of the old-age population in thatgiven age. For example, increasing the retirement age for men from age 60 to 61reduces the number of pensioners by the share of men age 60 in all men aged 60or older. In addition, it is assumed that about half of those for whom theretirement age increases would remain in the labour force, thus having an impacton employment. Changes in benefits (indexation or reductions) affect the benefitratio, and changes in beneficiaries affect the coverage ratio.

The importance of non-contributory pensionsto reduce poverty

Introduction and development of Georgia’s non-contributory pension system

Oneof themostsubstantialchallengesGeorgia facedduringthe transitiontodemocracyand a free-market economy in the mid-1990s was to reform its public welfare and




82

social security system inherited from the Soviet welfare regime.5 The foundation ofthe Soviet welfare and social security system was a guarantee of full employment.

The Georgian-Soviet public pension system provided differentiated pensions toretirees based on occupation. Pension benefits were financed on a pay-as-you-gobasis through the transfer of funds from state-owned enterprises to the localGeorgian department of the Soviet state insurance company, Gosstrakh. Toreceive a pension, women were required to have worked a minimum of 20 years,and men a minimum of 25 years (Gugushvili, 2009). In the late 1980s, coverageunder the old-age pension was nearly universal, paying between 60 per cent and100 per cent of the average wage in Georgia (Buckley, 1998).

Due to the demise of the Soviet Union in 1991, the newGeorgia had to reinvent itspension system, and Georgia faced substantial obstacles along the way arising fromeconomic and demographic circumstances. The economic profile of the early 1990swas typical for countries shifting from a centrally-planned to a free-market economy(Baumann, 2012). Dramatic hyperinflation and currency depreciation (1993–94),sharp economic decline and stagnation, and a state budget deficit throughout the1990s and into the early 2000s affected Georgia’s ability to keep up with pensioners’expectations.6 The difficult economic situation in the country was made worse bydemographic and social conditions: the population and birth rates declined,participation in the labour force fell, wages plummeted, and unemployment rose.Meanwhile, the informal economy grew and extensive external migration left behindan ageing population (DeCastello Branco, 1996, 1998;Wang, 1999; National Bank ofGeorgia, 2001; European Initiative – Liberal Academy Tbilisi, 2012).

By 1996, the government had replaced the Soviet system of differentiatedpensions, based on occupation, with a flat-rate pension system. Parametricadjustments meant that eligibility for pension benefits was available only forthose who had previously contributed to the system.7 Pension payments werelow and often delayed for months or even years because of budget limitations(European Initiative – Liberal Academy Tbilisi, 2012). More than a decade afterthe country’s independence (2004), pension payments were only 7 per cent ofGDP per capita (SSA, 2014). Pensioners had to rely on families, relatives, andneighbours, or turn to begging and street trading.

In 2006, the Georgian Parliament’s passage of the Law of Georgia on StatePension (2005) led to the closure of the existing pension system and the

5. In Georgia, as in other parts of the Soviet Union, the pay-as-you-go retirement system had been ineffect since the 1920s (EPRC, 2013).6. The “state budget” is Georgia’s central government’s budget.7. Pension benefits were determined in the following manner: “the number of those formallyemployed in the population was multiplied by the average salary and tax tariff, and then revenueswere divided by the number of pensioners. This meant that equal pensions were granted to all retireesregardless of their salaries during employment, length of service, or differences in pension type”(Gugushvili, 2009).




83

introduction of the non-contributory flat-rate basic pension, which had threecomponents: basic old-age pension, disability pension, and survivor pension, andwas disbursed when citizens reached retirement age (age 60 for women or age 65for men), or had an assessed disability, or for survivors following the death ofthe breadwinner. Most importantly, in 2006 the old-age pension became universalfor the elderly. Under reforms implemented in 2012, disability and survivorpensions have been separated into different social assistance programmes.However, in this study, in addition to the old-age pension, disability and survivorbenefits are also considered as part of the basic pension.

As mentioned, the basic pension was established as a flat-rate pension (2006),though it became a differentiated pension a year later. More specifically, about88 per cent of Georgian pensioners in 2006 received a standard minimum old-agepension of 38 Georgian Lari (GEL). For the remaining pensioners, including specialcategories of pensioners (veterans, widowed persons, politically repressed people,etc.), they were eligible for a higher old-age pension of between GEL 43 andGEL 129. In 2007, the old-age pension became differentiated following theintroduction of the seniority increment provided on the basis of years of work.As a result, only 33 per cent of pensioners (compared to 88 per cent in 2006)received a minimum old-age pension, which had increased to GEL 55 in 2007, whilethe rest of the pensioners became eligible for an entitlement of an additionalamount of GEL 2–10, based on the number of years of work. In 2012, the seniorityincrement was abolished. That same year, another change was implemented:different pensions were paid to pensioners of different ages. However, agedifferentiation was abolished by April 2013. More importantly – starting inSeptember 2013 – the basic old-age pension returned to a flat-rate pension.

The non-contributory pension system and poverty trends

Theestablishmentof thenon-contributorybasicpension in2006wasdrivenbytheneedto reduce Georgia’s substantial poverty rate. Since the implementation of the basicpension, the old-age pension programme – the basic pension system’s largestcomponent – has covered almost all the nation’s elderly and has supported theirincomes in old age by providing a minimum income floor for all. The old-age averagepension increased from 14.3 per cent of GDP per capita in 2006 to 26.5 per cent in2013 (see Figure 1).

At the same time, the relative elderly poverty rate decreased significantly: theshare of the population aged 60 or older who fell under 60 per cent of medianconsumption, decreased from 22.4 per cent in 2006 to 15.1 per cent in 2013.8

8. This article uses a relative poverty threshold equal to 60 per cent of median consumption, orGEL 130.20 per month (2013), according to the National Statistics Office of Georgia.




84

Figure 1. Old-age average pension as % of GDP per capita (2006–2013)

Sources: World Bank (2015c); National Bank of Georgia (2015); Pension Reform Unit of the Ministry of Economyand Sustainable Development of Georgia (2014); National Statistics Office of Georgia (2015); and authors’calculations.

Figure 2. Correlation between the increase in old-age average pension and thereduction in relative elderly poverty (2004–2013)

Sources: World Bank (2015c); National Bank of Georgia (2015); Pension Reform Unit of the Ministry of Economyand Sustainable Development of Georgia (2014); National Statistics Office of Georgia (2015); authors’ calculations.




85

Figure 2 shows the correlation between the increase in the old-age pension and thereduction in relative elderly poverty in the country from 2004–2013.

In Georgia in 2013, relative elderly poverty was substantially lower thanpoverty among other age groups. Figure 3 compares poverty among three groupsof the population: children (age 0–14), the working age population (age 15–59),and the elderly (aged 60 or older). In 2013, the relative elderly poverty ratewas 15.1 per cent, compared with 23 per cent relative poverty among theworking-age population and 25 per cent poverty among children.

The evidence suggests that the old-age pension reduces overall poverty inGeorgia. A UNICEF report (2012) has highlighted the impact of the old-agepension on consumption poverty in Georgia. The report shows that theold-age pension has a strong effect in reducing both relative and absolute extremeelderly, child, and household poverty, particularly for pensioners who live aloneor in small households.9 Moreover, the research findings suggest that if the pensionhad been removed in 2011, the percentage of pensioners who live below the relativepoverty threshold would have more than doubled (see Figures 4 and 5).

Additionally, Kidd and Gelders (2015) find that Georgia’s old-age pension iseffective in offering income security to households and children and in tacklingchild poverty at the macroeconomic level. The authors find that 46 per cent of

Figure 3. Relative poverty: Share of population (children, working age, elderly),under 60% of median consumption (2004–2013)

Source: National Statistics Office of Georgia (2015).

9. The UNICEF (2012) report used an absolute extreme poverty threshold: equivalent to GEL 71.70a month per adult, based upon the $1.25 benchmark used by the World Bank (identifying extremepoverty) and reflecting 2011 prices using the CPI.




86

households with children have at least one pensioner living in the household.Among chronically poor and highly vulnerable households with children,10

the old-age pension contributes around 20 per cent of household income. Infact, the old-age pension dramatically reduced child poverty in 2013 by69 per cent. The authors suggest that Georgia’s old-age pension is the mosteffective social pension among developing countries. Overall, the evidenceshows that the old-age pension has important potential for achievingGeorgia’s overall poverty reduction goals through its redistributive and socialimpacts.

The data from the Integrated Household Survey collected by theNational Statistics Office of Georgia (2015) show clearly the significance ofthe old-age pension in household income for different income groups. The averageold-age pension, as a share of total monthly income for households, nearly doubledbetween 2006 and 2013 in every income quintile except the lowest, which saw asmaller increase. However, the income from the old-age pension for households inthe lowest quintile is still crucial: the old-age pension contributed 34 per cent oftotal monthly household income in 2013 (Table 1).

10. By analyzing consumption dynamics, Kidd and Gelders (2015) categorized households withchildren into different economic classes.

Figure 4. Estimated effect of pension income on consumption poverty: % of households,children and pensioners living in extreme poverty before and after deduction of pensionincome (2011)

Source: UNICEF (2012).




87

It is worth noting that the relative elderly poverty rates for Georgia and theEuropean Union (EU) are comparable. The average at-risk-of-poverty rate inthe EU was about 14 per cent in 2013, which is comparable to the relativeelderly poverty rate of 15.1 per cent in Georgia for the same year.11 However,on an absolute basis, Georgia’s poverty remains higher than in the transitioneconomies of Europe and Central Asia. For instance, Schwarz et al. (2014)show that the percentage of all Georgian households that lived on $2.50 a daywas around 50 per cent of the population by the end of 2009. Furthermore,in 2009, Georgian households in poverty – those living on $2.50 a day – witha pensioner (aged 65+) in the household were less poor than householdswith pensioners who were older than age 80 and households with no pensioner(Schwarz et al., 2014). The World Bank (2015b) highlights that given theweaker buying power of many households in the European and Central

11. According to the glossary of the Statistical Office of the European Commission (Eurostat) “theat-risk-of-poverty rate” is the share of people with an equivalized disposable income (after socialtransfers) below the at-risk-of-poverty threshold, which is set at 60 per cent of the national medianequivalized disposable income after social transfers (Eurostat, 2014).

Figure 5. Estimated effect of pension income on consumption poverty: % of households,children and pensioners living in relative poverty before and after deduction of pensionincome (2011)

Source: UNICEF (2012).




88

Asian region, poverty rates are expected to rise in several oil-exporting andremittance-receiving countries.

Given the growing interest in social pensions, numerous studies document theirpoverty reduction effects.12 Additionally, most of the existing literature identifiestheir beneficial effect on a variety of outcomes, from expanding coverage toachieving vertical equity.

For example, Holzmann, Robalino and Takayama (2009) suggest thatnon-contributory pensions have a strong poverty reduction impact, particularly inmiddle-income countries where pensions can make up a significant share ofdifferent pension schemes. In most middle-income countries, dependence onnon-contributory pensions is caused by the expansion of substantial coverage gapsdue to the reduction in the number of workers who contribute to pensionschemes. In their discussion of non-contributory pensions in low-income

12. Social pensions, which are usually considered non-contributory cash transfers, differ from country tocountry; they can be universal (also referred to as basic pensions) or resource tested. Additionally, socialpensions may include first-tier public pensions, such as the minimum pension guarantees within mandatorycontributory pension schemes in European countries (Holzmann, Robalino, and Takayama, 2009).

Table 1. Basic pension spending and demographics (2009–2013)

2009 2010 2011 2012 2013

Pension spending (% of GDP)

Basic pension 4.1 3.8 3.5 4.1 4.7

Old-age pension 3.4 3.1 2.9 3.5 4.1

Survivor pensions/benefits 0.1 0.1 0.1 0.1 0.1

Disability pensions/benefits 0.6 0.6 0.5 0.5 0.5

Replacement rate (in % of average wage)

Average old-age pension 13.9 13.6 13.7 15.0 17.1

Demographics

Old-age pensioners 659,964 662,288 666,367 682,886 686,675

Recipients of survivor benefits 35,499 32,120 28,469 28,193 27,080

Recipients of disability benefits 139,932 138,614 129,599 122,056 122,940

Share of old-age pensioners in total numberof recipients of the basic pension (%)

78.7 79.2 80.6 82.0 82.1

Share of old-age pensioners in total population (%) 15.1 15.0 15.0 15.2 15.3

Old-age dependency ratio (65+/15-64) (%) 21.0 20.3 20.0 19.7 20.0

Sources: SSA (2014, 2015); Pension Reform Unit of the Ministry of Economy and Sustainable Development ofGeorgia (2014); National Statistics Office of Georgia (2014b, 2015); authors’ calculations.




89

countries, the authors found that the share of poor households with a pensionermember is low in low-income countries with younger populations. As such, socialpensions in those countries play some role in poverty mitigation, but the impact islimited in scope. In order to generalize, the authors suggest that social pensions inmost developing countries may represent core antipoverty programmes.

Additionally, research by Barrientos et al. (2003) on Brazil and South Africahighlights that non-contributory pensions have significant poverty reductionand prevention effects: they reduce household poverty and vulnerability andpromote the “functionings” of older people.13 Figari, Matsaganis and Sutherland(2013) find that there is a strong correlation between the resources allocated tominimum pension schemes and the reduction in poverty risk among the elderlyin the EU. Their view is that generous minimum pension schemes are moreeffective in reducing elderly poverty rates than the other pillars of pensionsystems.

The retirement income landscape in Georgia

Currently, the Georgian retirement income landscape includes income from thenon-contributory old-age pension, state compensation, private voluntarydefined-contribution pension schemes, and other forms of retirement incomesupport, such as formal health care, traditional family support, home and landownership, etc. However, the old-age pension is the largest system of publicsupport for the elderly and is fundamental in terms of its coverage, spending,and its contribution to elderly consumption. The other components of theGeorgian pension system, such as state compensation and voluntary pensions,have low coverage. Additionally, voluntary pensions in Georgia are not yet welldeveloped. Each of these sources of retirement income is discussed in more detail.

The basic pension

In terms of coverage, spending and poverty reduction, the basic pension is the mainpillar of the Georgian pension system and the largest source of public support forthe elderly. The largest of the basic pension programmes is the old-age pension; itis universal, flat-rate and financed by general budget revenues. Currently, the old-age pension is provided to all those reaching retirement age (age 60 for women orage 65 for men). Table 1 provides details on spending and other characteristics ofthe basic pension across 2009–2013, including the old-age pension, as well assurvivor and disability pensions. The table shows that in 2013 the Georgian basicpension was dominated by the “old-age” pension (86 per cent of all basic pensions).

13. Barrientos et al. (2003) define “functionings” as “the beings and doings that people value”.




90

The remainder was accounted for by survivor benefits (about 2 per cent of all basicpensions) and disability benefits (about 12 per cent).

During 2012–2013, the Georgian government raised the basic pension’s benefitlevels, which increased its funding levels to about 4.7 per cent of GDP. As of 2013,the replacement rate of the old-age pension reached 17 per cent of the average wage(see Table 1). Old-age pensioners represent 15 per cent of Georgia’s totalpopulation.14 By the end of 2013, the ratio of the population aged 65 or older tothe population aged 15–64 was over 20 per cent (see Table 1).

Basic pension spending and the government’s fiscal position

Since 2012, the Georgian government has increased spending on social provisions,including on the basic pension, state compensation, social assistance programmes,health care, etc.15 In 2013, total spending on social provisions accounted for about29 per cent of the state budget’s total expenditure (see Table 2) and reachedabout 8 per cent of GDP (see Table 3). The largest component of spending onsocial provisions is the basic pension (old-age, disability and survivors), whichaccounted for 60 per cent of total budget expenditure on social provisions in 2013and amounted to 4.7 per cent of GDP. The primary factor in this increase was theold-age pension, which accounted for over 15 per cent of the total state budget’sexpenditures and 4.1 per cent of GDP in 2013 (see Tables 2 and 3).

Given the ageing profile of Georgia, spending 4.1 per cent of GDP on old-agepensions is not unreasonable. However, such a relatively large commitment ofresources raises the cost of social provisions and challenges the sustainabilityof the basic pension programme. It is worth mentioning that Georgia’s publicexpenditure on non-contributory old-age pensions is in line with spending onnon-contributory pensions in New Zealand (3.9 per cent of GDP) but is lessthan in Denmark and the Netherlands (5.8 and 6.5 per cent of GDP, respectively),which have the highest costs for social pensions as a share of GDP in thehigh-income economies. However the level of pension spending in Georgia issignificantly higher than in other low- and middle-income economies where publicexpenditures on non-contributory pensions, in most cases, are at least 1 per cent ofGDP (HelpAge International, 2017).

The World Bank (2014) highlighted that the effect of the recent increase insocial spending in Georgia would be reflected in a deterioration of the fiscal deficitto 3.7 per cent of GDP in 2014 (see Table 3). The World Bank also underscoredthat high social spending might increase budget rigidity and restrict fiscal flexibility

14. In 2014, the share of the population aged 65 or older was 15.3 per cent in Georgia. This put Georgiaamong the oldest countries in the world (CIA, 2015).15. Social provisions in Georgia are to a significant degree financed from state budget general revenues.




91

in the future, which may make Georgia’s public finances more vulnerable.Additionally emphasized was the importance of further fiscal consolidation efforts(to create fiscal buffers) and the efficient management of public finances for thesustainability of medium- to long-term economic growth and job creation.

Alternatively, it is well established that social policy can potentially have apositive impact on the longer-term economic and fiscal situation of a country byreducing poverty and inequality. In this sense, cutting social spending may havenegative impacts on the longer-term fiscal situation, if the cuts weaken Georgia’seconomy. In fact, the Georgian government has prioritized social expendituresand has sought to maintain these at high levels to address the population’s need

Table 2. State budget (selected expenditure lines) (2009–2014)

Year 2009 2010 2011 2012 2013 2014

State budget total expenditure(billion GEL) 6.1 6.3 6.7 7.2 7.2 7.4*

Social provision (billion GEL) 1.4 1.5 1.5 1.7 2.0 2.5*

Share of social provision in state budget total expenditure (%) 23.2 23.4 23.1 23.8 28.8 33.7*

Basic pension (billion GEL) 0.8 0.9 0.9 1.0 1.3 1.5

Share of basic pension in state budget total expenditure (%) 13.3 13.4 13.5 14.1 17.3 19.6*

Old-age pension (billion GEL) 0.6 0.6 0.7 0.9 1.1 1.3

Share of old-age pension in state budget total expenditure (%) 9.9 10.1 10.4 12.7 15.1 17.0*

Notes: *2014 projection. Data is reported in current prices. Additionally, the average exchange rate of the USD toGEL during 2014 was 1.77. Figures in billion GEL are rounded.

Sources: MoF (2015); National Statistics Office of Georgia (2014b, 2015); SSA (2014, 2015); Pension Reform Unitof the Ministry of Economy and Sustainable Development of Georgia (2014); authors’ calculations.

Table 3. State budget (selected expenditure lines) and overall fiscal deficit, in % of GDP(2009–2014)

Year 2009 2010 2011 2012 2013 2014

State budget total expenditure 34.0 30.6 27.5 27.4 27.0 25.0*

Social provision 7.9 7.1 6.3 6.5 7.8 8.5*

Basic pension 4.5 4.1 3.7 3.9 4.7 5.0*

Old-age pension 3.4 3.1 2.9 3.5 4.1 4.3*

Overall fiscal deficit �9.2 �6.5 �3.5 �2.8 �2.5 �3.7**

Notes: *Projection for 2014 (Authors’ calculations); **Projection for 2014 (World Bank).

Sources: MoF (2015); National Statistics Office of Georgia (2014b, 2015); SSA (2015); World Bank (2014); authors’calculations.




92

for social assistance and health care (Government of Georgia, 2014b). Additionally,the Georgian government increased the basic pension up to GEL 160 a month(from GEL 150) starting in September 2015 (The Law of Georgia on State Budget,2015). In December 2016, the level of the basic pension was GEL 180 a month.

The pension system’s other components and other forms of retirement income

State compensation. State compensation was first introduced in 2006. Statecompensation aims to secure constitutional social protection for state andcivil workers (SSA, 2013). As a specific type of public transfer payment for civilservice that is financed from general revenues, the value of state compensation isset according to the importance of the work and services that specific workersperform for the state and civil society (The Law of Georgia on StateCompensation and State Academic Scholarship, 2005).16 As of 2013, totalexpenditures on state compensation reached about 0.2 per cent of GDP, coveringabout 18,012 people, or 3 per cent of the population aged 65 or older. Statecompensation provides a relatively generous retirement income: by the end of2013 its average benefit reached 58.3 per cent of GDP per capita.

Voluntary private pension schemes. The Georgian voluntary defined contributionprivate pension schemes – the third pillar – were implemented in 2007, institutedby insurance companies and offered by employers. By law, 80 per cent of thedefined contribution schemes’ investment reserves should be invested in Georgia. Atthe end of 2013, the number of participants amounted to 18,909 (about 0.9 per centof the labour force). In 2013, the accumulation of these schemes was only 0.048 percent of GDP. The Georgian voluntary pension schemes provide unfettered access topension savings, which can be drawn down for any purpose by their participants(USAID, 2011). As such, the voluntary pension schemes are more akin to regulatedsavings products than to pension products (Gadbury and Schou-Zibell, 2011).Also, there are no tax incentives on private pension savings in Georgia.

Other forms of retirement income support. The most common source ofretirement income support comes in the form of informal traditional family

16. State compensation is provided to: staff with rank and senior composition of military servicetransferred to the reserve of the military organs, internal affairs, special services of state security andGeorgian intelligence services; employees dismissed from the investigative service of the Ministry ofFinance of Georgia; staff dismissed from the public prosecutor’s office; members of the Parliamentof Georgia; officials of the Parliament of Georgia; members of the Constitutional Court of Georgia;judges of the Supreme Court of Georgia; judges of the Common Court of Georgia; members of thefamilies of former high political officials of Georgia; people with senior diplomatic rank; the chairpersonof the Chamber of Control and his/her deputy (deputies); servants of civil aviation (SSA, 2013).




93

support, as Georgians have a very strong sense of family and elderly careobligations. Among other sources of retirement income support are formalhealth care, access to individual financial assets (earnings from leasing property,dividends, etc.), financial support from family members (such as remittances andgifts), and non-financial assets such as homeownership and land.17

The basic pension outlook

Population ageing projections

The population of Georgia is ageing due to increasing longevity, below-replacementrate fertility and migration trends.18

• The ratio of population aged 60 or older to the population aged 15–59 isprojected to increase from 33 per cent in 2015 to 60 per cent by 2075 (see Figure 6).• Fertility rates are projected to remain quite low over the coming decades.Fertility rates in Georgia from 2010–2015 stood at around 1.81 children perwoman. Fertility rates are projected to increase slightly to 1.87 by 2070–2075.• Life expectancy at birth is currently 70.5 years for men and 77.7 years for women(2010–2015). Life expectancy for men at age 60 is projected to increase fromthe current 17.50 years to 22.93 years by 2075, and for women at age 60, from thecurrent 21.62 years to 26.73 years by 2075.

Basic pension spending projections

The most practical and politically appropriate assumption is that spending on thenon-contributory basic pension will increase in line with wage growth.Historically, the basic pension has experienced several ad hoc increases based onpolitical decisions. For instance, in 2012 and 2013, the basic pension increasedfaster than wages (World Bank, 2014). It is more predictable – and morepolitically feasible – that the government will keep the basic pension from fallingor may even increase it. Taking this into consideration, wage indexation wouldbe assumed to be the most appropriate policy response. Currently there are nopension indexation rules in Georgia. Figure 7 shows the baseline projection forthe basic pension, if wage indexation were implemented. Figure 8 shows thebasic pension projection under current law and therefore does not assume anindexation policy.

17. It is worth noting that during 2012–2013 the Georgian government introduced universal healthcare, resulting in an increase in the number of out-patient visits (National Statistics Office of Georgia,2014b).18. This is based on United Nations population data and projections (ESA, 2015).




94

Figure 6. Ratio of population older than age 60 to population 15–59 (2015–2075)

Notes: The old-age dependency ratio drops from 64% to 55% in 2055–2065, due to rapid declines in fertility rates inthe early 2000s and associated small population cohorts (ESA, 2105).

Source: ESA (2015).

Figure 7. Baseline projection: Average benefit as % of GDP per worker, basic pensionspending as % of GDP (2015–2075)

Source: Authors’ calculations.




95

The baseline projection shows that indexing the basic pension to wage growthnot only helps to maintain its adequacy, but would be important to thecontinued alleviation of poverty in Georgia. However, this baseline projectionalso shows that the long-term cost of the basic pension system increases from4.7 per cent of GDP in 2015 to 7.4 per cent of GDP by 2075 (see Figure 7). Thisincrease, which is in line with an increase in the old-age dependency ratio, mayjeopardize the sustainability of the basic pension system over the long term.

To maintain the system’s sustainability without introducing any other reforms,it will be necessary to raise taxes to finance the basic pension out of general taxrevenues or to cut benefits. None of these options is appropriate. Fiscal resourcesin Georgia are scarce – taxes are already high, and there is a constitutionalprovision that bars the government from increasing taxes (World Bank, 2014).Alternatively, reductions in benefit levels would undermine the poverty alleviationobjective of the basic pension. An additional option is to cut state budgetexpenditures on some other governmental programmes, though this is notpolitically feasible.

Options for reform

Two factors limit the future sustainability of Georgia’s non-contributory basicpension. The country is characterized by an ageing population and scarce public

Figure 8. Projection, based on the current legal conditions: Average benefit as % ofGDP per worker, basic pension spending as % of GDP (2015–2075)





96

financial resources, which create many pressing needs for social provisions.Consequently, the key goals of the basic pension reform are how to:• Maintain reasonable funding burdens while also maintaining the long-termsustainability and adequacy of the pension system; and• Ensure sufficient flexibility in the system to adapt to the evolving demographicenvironment.

In the following sections, four pension reform scenarios are considered ingreater detail.

Scenario 1: Indexing future pension benefits to prices

Among the different parametric reform options, indexation to prices would beconsidered to protect the purchasing power of pensions. It is worth mentioningthat the Georgian government intends to link the basic pension to inflation until2020 (Government of Georgia, 2014b).

Figure 9 shows the impact of the price indexation reform on the basicpension: it would result in low pension spending (1.9 per cent of GDP in2075), but also in lower, inadequate pension benefits (the average benefit fallsfrom 12.5 per cent of GDP per worker to 3.2 per cent by the end of the

Figure 9. Changes in price indexation, relative to baseline: Average benefit as % ofGDP per worker, basic pension spending as % of GDP (2015–2075)





97

projected period). The lower benefits will undermine the ability of the basicpension to alleviate relative poverty.

Scenario 2: Increasing retirement ages

Generally, a parametric pension reform, which aims to reduce the cost of the pensionsystem, may include curtailing eligibility (i.e. changing the number of eligiblepensioners by increasing the statutory retirement ages). Curtailing eligibility inGeorgia should be considered, as life expectancy at age 60 is projected to increasein the coming decades. This option potentially can help to contain pension spending,avoid a socially and politically undesirable decrease in the pension replacement rate,encourage later exit from the labour market, contribute to economic growth, andimprove the fiscal sustainability of the pension system (Eich, Gust and Soto, 2012).

Table 4 proposes reasonable increases in Georgia’s retirement ages, starting in2025, and a schedule for the equalizing of retirement ages for men and women.Female life expectancy is greater than for men. The retirement ages can beincreased and equalized gradually: to age 65 for women after 2020, and then toage 68 for both men and women by 2060. Figure 10 shows the impact ongovernment pension expenditures and on pension benefit levels of changingand equalizing the retirement ages: basic pension spending increases from4.7 per cent of GDP in 2015 to 5 per cent in 2075.

This policy reform option also has the potential to stabilize the ratio ofpensioners to the working-age population (Figure 11). Additionally, this optionmay induce later exit from the labour market, which could contribute to helpconsumer spending and economic growth.

However, it should be noted that increasing and equalizing Georgia’s statutoryretirement ages require considerations that take into account the nationalenvironment: high levels of unemployment and a large informal economy. For

Table 4. Statutory retirement ages: Proposed increases (2015–2060)

Year Retirement age (Men) Retirement age (Women)

2015 65 60

2025 65 62

2030 65 65

2040 66 66

2050 67 67

2060 68 68

Source: Authors’ proposal.




98

Figure 10. Changes in retirement ages, relative to baseline: Average benefit as % ofGDP per worker, basic pension spending as % of GDP (2015–2075)


Figure 11. Changes in retirement ages, relative to baseline: Ratio of pensioners toworking-age population, beneficiaries per population 60+ (2015–2075)





99

instance, over 60 per cent of working-age Georgians were self-employed in 2013.Self-employment is most common among rural populations (National StatisticsOffice of Georgia, 2014a). Increasing and equalizing statutory retirement ageswill be easier if the Georgian agricultural sector develops – this is one of theGeorgian government’s highest priorities (Government of Georgia, 2015).Considerable recent and proposed increases in the financial resources flowing tothe agricultural sector, as well as improved government strategies, can result in thesector’s medium- and long-term development (The Law of Georgia on StateBudget, 2013, 2014 and 2015). However, if statutory retirement ages increase,the challenge of low employment rates in urban areas may lead to an increasein urban poverty levels. This will need to be addressed by policy-makers, andpolicies aimed at increasing employment rates in Georgia’s urban areas shouldbe implemented.

The differences in life expectancy among socio-economic groups is anotherfactor to consider. The International Actuarial Association (IAA) highlights thatlongevity, and particularly healthier longer lives, are greater for people in highersocio-economic groups. The IAA also suggests that the effect of increasing thepension eligibility age in line with increases in life expectancy will have a largerrelative impact on those in lower socio-economic groups who rely most on socialsecurity benefits (IAA, 2016). In considering the effects on Georgia’s lowersocio-economic groups of increasing eligibility ages in the basic pension,policy-makers need to reconsider supplementary programmes, including thedelivery of effective social assistance and health care services targeted ondisadvantaged groups.

Additionally, so as to prevent increases in Georgia’s poverty rate, an increase inthe retirement ages should be supplemented with other social assistance measuresthat protect the incomes of those who cannot continue to work due to significanthealth issues.

Increasing and equalizing retirement ages may also impact Georgian womenunfavourably. Typically, women in Georgia tend to stay at home andemployment among women is low (National Statistics Office of Georgia,2014a). It should be politically feasible in Georgia to both facilitate women’slabour force participation and to limit early retirement for men and women.Moreover, the potential supply of, and demand for, older workers shouldbe explored, and policies to encourage employment of older people should beimplemented.

The above mentioned considerations with regard to increasing and equalizingstatutory retirement ages in Georgia are not intended to indicate that this scenariois any more or less favoured than others, but rather that its consequences must beunderstood and that appropriate actions are needed to off-set any adverseconsequences, while taking advantage of the positive effects.




100

In summary, in response to fiscal stress, Georgia will need to gradually increasethe age of eligibility for the basic pension with a relatively long lead time thatenables policy-makers to implement policies that will promote growth andemployment, and protect the most vulnerable populations. This long lead timewill also enable individuals to adjust their behaviour, especially with respect toemployment patterns and retirement planning (particularly for women).

Scenario 3: Reducing pension generosity through means testing

Reducing the generosity of benefits is another reform option that would reduce thefiscal burden of the pension system. It may be done by the adoption of meanstesting. Such a parametric change in the Georgian basic pension system can bedone simultaneously with the introduction of contributory private pensionschemes, which has been suggested by the Georgian government since 2012 andis reflected in government documents and laws (Government of Georgia, 2012,2013, 2014 and 2015; The Law of Georgia on State Budget, 2013, 2014 and2015). A recent government report (MoESD, 2016) suggests that mandatorysecond-pillar private defined contribution pension schemes should beimplemented in Georgia (MoESD, 2016).19

The adoption of means testing, and particularly imposing an ex-ante meanstest of income or assets, would reduce the fiscal cost of the basic pensionin Georgia. Willmore (2007) suggests that ex-ante means tests are common(e.g. in Australia, Chile, Costa Rica, India, South Africa, United States) andattractive for policy-makers, though their administrative costs are high and targetingis crude. Another option to be considered is the adoption of an ex-post incometax. However, despite the relative simplicity of administration, an income test canbe politically unpopular and is unlikely to materialize in Georgia.

Assuming that private defined contribution pension schemes functionsuccessfully, these will enable high-earning participants to make sufficientretirement savings so that they would no longer require a basic pension. Thebasic pension would exist primarily for poverty relief – and not to financeold-age support for people with higher earnings. This policy reform optionwould reduce participation in the basic pension system by 15 per cent (by meansof restricted eligibility criteria) starting in 2045, when pension payments will beavailable for the participants in the proposed defined contribution pensionschemes (see Figure 12).

19. Implementation details for the second pillar contributory pension schemes are not clearly describedin the government documents. The most recent government report states that “… soft mandatory,quasi-mandatory approach should apply” (MoESD, 2016).




101

However, Figure 13 shows that this proposal would not be sufficient alone tomaintain the basic pension system’s sustainability. Despite the fact that the costof funding the system – if a second pillar defined contribution system isadded – is lower than in the baseline projection, pension spending still increasesfrom 4.7 per cent of GDP in 2015 to 6.3 per cent of GDP by 2075 (Figure 13).

A transition to a multi-pillar pension system should be considered in Georgia.Second-pillar private defined contribution pension schemes are a realistic meansof generating savings within a society and could potentially play a role in economicdevelopment (Gadbury and Schou-Zibell, 2011). Nevertheless, there are severaluncertainties and challenges related to their implementation. The following arethe most relevant.

The development of a second-pillar private defined contribution pensionscheme that is mandatory may create uncertainty in Georgia: this systemicreform option may not be implemented, or, if implemented may be reversed orchanged. Despite the fact that implementing a multi-pillar pension system wasfirst considered as a suitable option for Georgia in 1999, and that pension systemreform has been debated for years, implementation of the reform has been delayeddue to political and economic circumstances and weak technical capacity(Gugushvili, 2009). However, Georgia has made huge strides in democratic andeconomic development over the past decade and is beginning to show significant

Figure 12. Reducing the generosity of benefits by means testing, relative to baseline:Beneficiaries per population aged 60+ (2015–2075)





102

potential to consider such a systemic pension reform. Moreover, the establishmentof a multi-pillar pension system is an overriding priority for the newly-electedGeorgian government.

Additionally, there is a level of uncertainty regarding whether a definedcontribution pension scheme can function well in Georgia, given that such a systemwould require developing and fostering a sound and efficient financial system and aregulatory framework.

Reducing the generosity of the basic pension benefit through means testingmight be challenging, because means testing can introduce distortions related todisincentives for saving for retirement (Biggs, 2011). As discussed in Barr andDiamond (2006), pension system changes would also distort the labour market,though “it is mistaken to focus on labour market distortions while ignoringor downplaying the contributions to the various goals of pension systems– contributions that are not available without distortions”. The authors suggestthat policy has to balance labour market efficiency against the various objectivesof pension systems. Moreover, research by Sørensen et al. (2016) highlights thatthe overall design of a pension system and its pillar interactions (including basic,defined contribution and defined benefit pensions, and the use of targeted

Figure 13. Reducing the generosity of benefits by means testing, relative to baseline:Average benefit as % of GDP per worker, basic pension spending as % of GDP(2015–2075)





103

benefits in particular) play crucial roles in mitigating risks and uncertainties relatedto private pensions, including concerns about any labour market distortions thatwould result.

The above-mentioned issues are important for the discussion when consideringsystemic pension reform and the adoption of means testing in Georgia’s basicpension system, and public policy decisions must aim at an optimal resolution ofthese conflicting considerations.

Scenario 4: Simultaneous reforms – Increasing retirement agesand reducing the generosity of benefits by means testing

Some of the previously discussed options for improving and developing the basicpension system in Georgia can be combined into a more appropriate policy reformchoice. The medium- and long-term sustainability of the basic pension system canbe achieved by undertaking a combination of means testing and a gradual increaseand equalization of the retirement ages. This reform mix assumes increasing thestatutory retirement age for women to age 65 after 2020, and for both men andwomen to age 68 by 2060; and introducing means testing in 2045, as the proposeddefined contribution pension schemes become fully functional.

Figure 14. Simultaneous reforms – Increasing retirement ages and reducing thegenerosity of benefits by means testing, relative to baseline: Average benefit as % ofGDP per worker, basic pension spending as % of GDP (2015–2075)





104

Figure 14 shows that a reform mix will help to ensure a sustainable fiscalposition: the cost of the basic pension will be stabilized at around 5 per cent ofGDP throughout the projected period. This policy reform option will also helpto maintain pension benefits on a consistent level (i.e. the average benefit isstabilized at 12 per cent of GDP per worker).

In sum, this reform mix strikes a relatively good balance between concerns forfiscal sustainability and benefit adequacy. It would also improve the basicpension system’s political viability.

Conclusion

The main objectives of Georgian social policy are to reduce poverty and mitigate thesocial risks caused by population ageing (Government of Georgia, 2015). A reformof the non-contributory basic pension system can play a substantial role in this.Therefore, the main goals of Georgian pension reform should be to maintain thesustainability of the basic pension system and to ensure its poverty alleviationrole. The appropriate reform mix might include an increase in retirement agesand a reduction in the generosity of benefits through means testing.

This article finds that increasing the statutory retirement ages for women to age65 after 2020, and then increasing the retirement ages for both men and women toage 68 by 2060 is an attractive policy option with the three-fold aim to ensure thatbasic pension benefits are adequate, to contain government spending on basicpensions, and potentially improve fiscal sustainability. This policy reform optionhas the further benefit of potentially stabilizing the ratio of pensioners to theworking-age population in Georgia, thus supporting higher labour forceparticipation and contributing to economic growth.

Means-tested benefits have the advantage of being less costly than Georgia’scurrent system of universal benefits. The most important implication associatedwith means testing – which should be addressed by policy-makers – is theinherent potential to provoke disincentives to work and save. This may adverselyaffect the development of the proposed second pillar defined contributionpension schemes. The main challenge related to this policy option in Georgiawould be to create policies that avoid the negative consequences of means testing,while making means testing an efficient solution for improving the sustainabilityof the basic pension system. Eventually, means testing would transform thenature of the basic pension: the universal basic pension would become a pensionfor the poor.

If the political will exists, then the reform of the non-contributory basic pensionsystem may be complemented by a reform of Georgia’s state compensation, as partof a broader pension system reform. Given that state compensation is offered to




105

special categories of state and civil workers – as a future promise to secureconstitutional social protection for state and civil workers – support for it ispolitically strong.

Economic growth and institutional improvement in Georgia may facilitate thegradual development of the proposed second pillar pension system, and eventhe voluntary third pillar. The generosity of the basic pension system might bereduced, contingent on the development of contributory pension schemes.

Finally, in the medium-term perspective, the basic pension programme must becomplemented by the implementation of more effective social assistance and healthcare programmes targeting the elderly as well as other vulnerable populationgroups in extreme poverty. This should enhance the effectiveness of the basicpension in mitigating poverty in Georgia.

Bibliography

Barr, N.; Diamond, P. 2006. “The economics of pensions”, in Oxford Review of Economic

Policy, Vol. 22. No. 1.

Barrientos, A. et al. 2003. Non-contributory pensions and poverty prevention: A comparative

study of Brazil and South Africa (Final report, DFID Project R7897: Pensions and poverty

prevention). London, HelpAge International.

Baumann, E. 2012. “Post-Soviet Georgia: It’s a long, long way to ‘modern’ social

protection”, in Economies et Sociétés (Séries: Développement, croissance et progrès),

Vol. 46, No. 2.

Biggs, A. G. 2011. “Means testing and its limits”, in National Affairs, Vol. 9, Fall.

Buckley, C. 1998. “Obligations and expectations: Renegotiating pensions in the Russian

Federation”, in Continuity and Change, Vol. 13, No. 2.

CIA. 2015. The world factbook. Washington, DC, Central Intelligence Agency.

Clements, B.; Eich, F.; Gupta, S. 2014. Equitable and sustainable pensions: Challenges and

experience. Washington, DC, International Monetary Fund.

De Castello Branco, M. 1996. Georgia: from hyperinflation to growth (IMF Survey).

Washington, DC, International Monetary Fund.

De Castello Branco, M. 1998. Pension reform in the Baltics, Russia and other countries of the

Former Soviet Union (BRO) (IMF working paper, No. 98/11). Washington, DC,

International Monetary Fund.

EPRC. 2013. Do we need a pension reform?. Tbilisi, Economic Policy Research Center.

Eich, F.; Gust, C.; Soto, M. 2012. Reforming the public pension system in the Russian

Federation (IMF working paper, No. 12/201). Washington, DC, International Monetary

Fund.

ESA. 2015. World population prospects: The 2015 revision. New York, NY, United Nations

Department of Economic and Social Affairs – Population Division.




106

http://www.helpage.org/silo/files/noncontributory-pensions-and-poverty-prevention-a-comparative-study-of-brazil-and-south-africa-.pdf

http://www.helpage.org/silo/files/noncontributory-pensions-and-poverty-prevention-a-comparative-study-of-brazil-and-south-africa-.pdf

http://horizon.documentation.ird.fr/exl-doc/pleins_textes/divers12-07/010056159.pdf

http://horizon.documentation.ird.fr/exl-doc/pleins_textes/divers12-07/010056159.pdf

http://www.nationalaffairs.com/doclib/20110919_Biggs.pdf

https://www.cia.gov/library/publications/the-world-factbook/fields/2010.html#4

http://www.elibrary.imf.org/view/IMF071/20751-9781616359508/20751-9781616359508/20751-9781616359508.xml

http://www.elibrary.imf.org/view/IMF071/20751-9781616359508/20751-9781616359508/20751-9781616359508.xml

https://www.imf.org/external/pubs/ft/survey/pdf/092396b.pdf

http://www.imf.org/external/pubs/ft/wp/wp9811.pdf

http://www.imf.org/external/pubs/ft/wp/wp9811.pdf

https://www.imf.org/external/pubs/ft/wp/2012/wp12201.pdf

https://www.imf.org/external/pubs/ft/wp/2012/wp12201.pdf

https://esa.un.org/unpd/wpp/

European Initiative – Liberal Academy Tbilisi. 2012. საქართველოს ეკონომიკური

ტრანსფორმაცია: დამოუკიდებლობის 20 წელი [The economic transformation

of Georgia: 20 years of independence] (Interim report). Tbilisi.

Eurostat. 2014. Glossary: At-risk-of-poverty rate (Statistics explained). Luxembourg,

Statistical Office of the European Commission.

Figari, F.; Matsaganis, M.; Sutherland, H. 2013. “The financial well-being of older people

in Europe and the redistributive effects of minimum pension schemes”, in Rivista Italiana

Degli Economisti, Vol. 18, No. 2.

Gadbury, J.; Schou-Zibell, L. 2011. Pensions and pension reform in Georgia (ADB Central

and West Asia Working Paper No.1). Metro Manila, Asian Development Bank.

Government of Georgia. 2012, 2013, 2014a, 2015. Government programme: For strong,

democratic and unified Georgia. Tbilisi. [in Georgian].

Government of Georgia. 2014b. Social-economic development strategy of Georgia:

Georgia 2020. Tbilisi.

Gugushvili, A. 2009. “Political economy of old-age pension reforms in Georgia”, in

Caucasian Review of International Affairs, Vol. 3, No. 4.

HelpAge International. 2017. Pension watch: Social pensions database. London.

Holzmann, R.; Hinz, R. 2005. Old Age Income support in the 21st Century. Washington,

DC, World Bank.

Holzmann, R.; Robalino, D. A.; Takayama, N. 2009. Closing the coverage gap: The

role of social pensions and other retirement income transfers. Washington, DC,

World Bank.

Holzmann, R.; Hinz, R.; Dorfman, M. 2008. Pension systems and reform conceptual

framework (Discussion paper, No. 0824). Washington, DC, World Bank.

IAA. 2016.Determination of retirement and eligibility ages: Actuarial, social, and economic impacts.

Ottawa, International Actuarial Association.

IMF. 2011. The challenge of public pension reform in advanced and emerging economies.

Washington, DC, International Monetary Fund.

Kidd, S.; Gelders, B. 2015. Child wellbeing and social security in Georgia: The case for moving

to a more inclusive national social security system. Tbilisi, UNICEF Georgia.

MoESD. 2016. Georgian pension reform: Reform of the universal pension benefit and

introduction of the supplementary pension scheme. Tbilisi, Ministry of Economy and

Sustainable Development of Georgia.

MoF. 2015. Basic economic and financial indicators. Tbilisi, Ministry of Finance.

MoF. various years. Country basic data and directions (2012–2015, 2013–2016, 2014–2017,

2015–2018). Tbilisi, Ministry of Finance.

National Bank of Georgia. 2001. Bulletin of Monetary and Banking Statistics, Vol. 12,

No. 34. [in Georgian].

National Statistics Office of Georgia. 2014a. Employment and unemployment 2013

(Annual report). Tbilisi. [in Georgian].

National Statistics Office of Georgia. 2014b. Statistical yearbook of Georgia 2013. Tbilisi.




107

http://www.ei-lat.ge/images/stories/The_Economic_Transformation_of_Georgia_-_20_Years_of_Independence_Interim_Report_geo.pdf

http://www.ei-lat.ge/images/stories/The_Economic_Transformation_of_Georgia_-_20_Years_of_Independence_Interim_Report_geo.pdf

http://ec.europa.eu/eurostat/statistics-explained/index.php/Glossary:At-risk-of-poverty_rate

http://www.adb.org/sites/default/files/publication/29098/cwrd-wp01.pdf

http://gov.ge/index.php?lang_id=GEO&sec_id=68

http://gov.ge/index.php?lang_id=GEO&sec_id=68

http://www.adb.org/sites/default/files/linked-documents/cps-geo-2014-2018-sd-01.pdf

http://www.adb.org/sites/default/files/linked-documents/cps-geo-2014-2018-sd-01.pdf

http://www.pension-watch.net/about-social-pensions/about-social-pensions/social-pensions-database/

http://siteresources.worldbank.org/SOCIALPROTECTION/Resources/SP-Discussion-papers/Pensions-DP/0824.pdf

http://siteresources.worldbank.org/SOCIALPROTECTION/Resources/SP-Discussion-papers/Pensions-DP/0824.pdf

http://www.actuaries.org/LIBRARY/Papers/PIWG_Retirement_and_Eligibility_Ages_Paper_final_11March2016.pdf

http://www.imf.org/external/np/pp/eng/2011/122811.pdf

http://socialprotection-humanrights.org/wp-content/uploads/2016/04/UNICEF_Child_Wellbeing_ENG_.pdf

http://socialprotection-humanrights.org/wp-content/uploads/2016/04/UNICEF_Child_Wellbeing_ENG_.pdf

http://economy.ge/uploads/meniu_publikaciebi/ouer/Goverment_Report_on_pensions_Final_ENG.pdf

http://economy.ge/uploads/meniu_publikaciebi/ouer/Goverment_Report_on_pensions_Final_ENG.pdf

http://www.mof.ge/en/4541

http://www.mof.ge/en/4543

https://www.nbg.gov.ge/uploads/publications/bulletinstatistics/statbiul/2013/annualgeo/bulletin2001geo.pdf

http://www.geostat.ge/cms/site_images/_files/georgian/labour/dasaqmeba%20da%20umuSevroba%202014%20pres-relizi.pdf

http://www.geostat.ge/cms/site_images/_files/yearbook/Yearbook_2014.pdf

National Statistics Office of Georgia. 2015. Data. Tbilisi, Georgia.

Palacios, R.; Knox-Vydmanov, C. 2014. “The growing role of social pensions: History,

taxonomy and key performance indicators”, in Public Administration and Development,

Vol. 34, No. 4.

Pension Reform Unit. 2014. Data. Tbilisi, Ministry of Economy and Sustainable

Development.

Schwarz, A. et al. 2014. The inverting pyramid: Pension systems facing demographic challenges

in Europe and Central Asia (Europe and Central Asia Reports). Washington, DC,

World Bank.

Sørensen, O. B. et al. 2016. “The interaction of pillars in multi-pillar pension systems: A

comparison of Canada, Denmark, Netherlands and Sweden”, in International Social

Security Review, Vol. 69, No. 2.

SSA. 2013. State compensation: Whom the state compensation may be assigned to. Tbilisi,

Social Service Agency.

SSA. 2014. Social statistics 2013: Annual report. Tbilisi, Social Service Agency. [in Georgian].

SSA. 2015. Social statistics. Tbilisi, Social Service Agency.

UNICEF. 2012. Georgia: Reducing child poverty – A discussion paper. Tbilisi,

UNICEF Georgia.

USAID. 2011. Pension development in Georgia (Final report). Tbilisi, United States Agency

for International Development.

Willmore, L. 2007. “Universal pensions for developing countries”, in World Development,

Vol. 35, No. 1.

Wang, J.–Y. 1999. The Georgian hyperinflation and stabilization (IMF working paper,

No. 99/65). Washington, DC, International Monetary Fund.

World Bank. 2014. Georgia: Public expenditures review – Strategic issues and reform agenda

(Report, No. 78143-GE 2014). Washington, DC.

World Bank. 2015a. Data: GNI Per Capita, PPP. Washington, DC.

World Bank. 2015b. Europe and Central Asia overview. Washington, DC.

World Bank. 2015c. GDP per capita (Current US$). Washington, DC.

Legislation of the Government of Georgia

Law on State Compensation and State Academic Scholarships (No. 2549 of

27 December 2005)

Law on State Pensions (No. 2442 of 23 December 2005)




108

http://www.worldbank.org/content/dam/Worldbank/Feature%20Story/ECA/ECA-Pensions-Report-2014.pdf

http://www.worldbank.org/content/dam/Worldbank/Feature%20Story/ECA/ECA-Pensions-Report-2014.pdf

http://ssa.gov.ge/index.php?sec_id=24&lang_id=ENG

http://ssa.gov.ge/files/01_GEO/PUBLIKACIEBI/FILES/STAT-2013.pdf

http://www.unicef.org/ceecis/UNICEF_Child_PovertyENG_web_with_names.pdf

http://pdf.usaid.gov/pdf_docs/pnadz449.pdf

http://www.imf.org/external/pubs/ft/wp/1999/wp9965.pdf

http://www-wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2014/07/31/000333037_20140731104244/Rendered/PDF/781430GE0v10RE0Box0385291B00PUBLIC0.pdf

http://data.worldbank.org/indicator/NY.GNP.PCAP.PP.CD

http://www.worldbank.org/en/region/eca/overview#1

http://data.worldbank.org/indicator/NY.GDP.PCAP.CD

http://www.ilo.org/dyn/natlex/natlex4.detail?p_lang=en&p_isn=98946&p_classification=15.05

http://www.ilo.org/dyn/natlex/natlex4.detail?p_lang=en&p_isn=88317

I

R

S

S

E V I E W

E C U R I T Y

O C I A L

NTERNATIONAL

B O O K SB O O K S

BOOK REVIEW

OECD. Pensions at a glance 2015: OECD and G20 indicators. Paris, Organisation for

Economic Co-operation and Development. 2015. 374 pp. ISBN 978-92-64-24063-6.

The sixth OECD bi-annual overview of pension systems and policies coincides with the tenthanniversary of this series, which has been fostering pension reforms, assessing outcomes andevaluating their future potential for ensuring broad coverage and adequacy, reducing genderinequality in access, and, more generally, preventing old-age poverty, while maintaining the financialsustainability of public pensions (pay-as-you-go) among its 34 member states and G20 countries.This issue’s particular concern is for pension policy’s greater focus on social sustainability.

The report consists of 11 chapters, with the first three providing, respectively, a cross country(OECD and G20) comparative analysis of the most recent pension reforms (Chapter 1), the role offirst-tier pensions (Chapter 2) and the effect of incomplete careers on pension entitlements (Chapter3). Continuing with past practice, the seven chapters that follow provide extensive fact sheets andtables, each devoted to specific components and indicators of pension reforms, focusing,respectively, on: the sensitivity of replacement rates to the pension models’ parameters (Chapter 4);the design of national pension systems (Chapter 5); the methodology and assumptions underpinningpension entitlements (Chapter 6); demographic and labour market contexts (Chapter 7); incidence ofincomes and poverty among older people (Chapter 8); the various components of retirement incomesystems (Chapter 9); and the structure, coverage and allocation of private pensions assets and ofpublic pension reserve funds (Chapter 10). The volume concludes with individual country profiles(Chapter 11).

The report starts with an overview of recent pension reform achievements and shortcomings(Chapter 1). It notes that most governments have reformed their pension systems, improving therebytheir financial sustainability, while also ensuring good living standards for some groups of pensioners(especially men who had stable, long-term jobs) and strengthening safety nets for vulnerable groupsof pensioners. However, it points to growing risks on the horizon. These include the impact of theglobal financial crisis and the subsequent sluggish economic growth, and large and growing publicdeficits and debt (the latter rising from 55 per cent of GDP in 2007 to 88 per cent in 2014 onaverage across OECD countries), which limit governments’ spending capacity. Further challengesare rapid demographic ageing and the changing world of work – characterized by high youth andelderly unemployment, late entry and relatively early departure from the labour market, discontinuedcareers, and the growing incidence of precarious job status. Taken together, these developmentsdirectly affect the capacity of the workforce to earn enough to save for retirement (e.g. interruptedand irregular job duration and status, poor pay and few career prospects for the less educated orless qualified, family responsibilities, etc.). Unsurprisingly, this trend has given rise to a new concernthat has resulted from the gradual shift – since the mid-1980s – of relative income poverty rates fromamong the elderly towards the young, thus endangering their capacity to contribute to pensionsavings and putting at risk the sustainability of both current and future pensions.



109

http://doi.org/info:x-wiley/isbn/9789264240636

http://doi.org/info:x-wiley/isbn/9789264240636

To address the challenge of financial sustainability, many OECD countries introduced reforms tolimit future pension expenditure until 2050, i.e. by strengthening incentives to encourage individualsto extend their working lives by increasing the minimum and/or the official retirement age, graduallyeliminating early retirement options, changing the indexation of pension benefits to prices ratherthan to wages (which tend to rise faster than prices), resulting in pension value – compared toearnings – declining over time. So, the longer retirement lasts, the more important indexationbecomes for adequacy. It is, however, noteworthy that practically no country resorted to directbenefit cuts, using instead narrower targeting or adjusting initial benefits to new retirees, orincreasing taxes and contributions in defined benefit (DB) schemes (without granting additionalpension entitlement), lowering tax deductions on pension contributions or on pension assets (thoughsuch measures can discourage participation and result in lower savings in voluntary schemes).Arguably, such measures may help government to cope with budgetary constraints, but they alsotend to increase pensioner poverty.

To improve pension adequacy, several governments increased coverage or benefits levels or both.In schemes with a direct link between contributions paid and benefits received in retirement, highercontributions or longer periods of contribution were introduced, especially for defined contribution(DC) schemes to improve the level of benefits (notably in Australia, Israel, Norway and the UnitedKingdom), and for voluntary private pension schemes. Some governments granted retroactive pensioncredits to alleviate the impact of missing contribution years on pension levels or reduced the effectivetaxation of pensioners’ income. Other measures included lowering the administrative managementcosts of schemes and strengthening measures to improve the security of pension investments, byincreasing investor choice for funded schemes, or improving governance and risk management ofpension plans. For example, Chile, introduced minimum and maximum limits for foreign currencyhedges to reduce risk. Ireland introduced a new benefit security in case of company bankruptcy,increasing risk reserves from 2016 and imposing stricter reporting of actuarial reserves and theage-dependent capitalisation amount for defined benefits (since 2014).

As regards pension benefit coverage, many OECD countries have recently extended coverage togroups not previously covered, while others introduced new benefits to extend coverage (for example,Japan has decreased, from April 2017, the qualifying period for national pensions from 25 to 10 years,benefiting short career workers, while pension insurance has been extended to more part timers sinceOctober 2016). But extended coverage is difficult to achieve even in mandatory schemes in countrieswith a large informal economy (Mexico, Chile and some G20 countries). A number of countries alsoimproved pension benefit coverage by targeting vulnerable groups.

However, huge cross-country differences remain in the so-called replacement rate, which is onemeasure of adequacy (net value being equal to the ratio of net pension entitlement to life-timeaverage net earnings). The net value across the OECD for average-income earners is 63 per cent,but varies from 28 per cent in Mexico to 105 per cent in Turkey. Low-income earners generallyhave higher net replacement rates in most OECD countries, for which the average equals 75 per cent.

The major challenges ahead are: slow economic growth; demographic ageing; labour marketshortcomings (unemployment, precarious and low paying jobs, lack of skill upgrading opportunities,discontinued careers); large increases in government indebtedness; and low interest rates, whichlimit the capacity of pension funds, life insurance companies and pension savers to achieveadequate pension levels. In short, pension systems might maintain inequalities that appeared duringthe working life by not taking into account disparities in mortality rates that are correlated with



110

poverty, which makes them less distributive, and might risk, according to the authors, to make some ofthem regressive.

Chapter 2 looks successively into the design of first-tier pensions (basic, minimum and means-tested old-age social assistance payments), addressing their structure, eligibility rules, the level ofbenefits as a proportion of average earning versus the old-age poverty rate, as well as other formsof assistance (including rent or health), while also analysing the implications of indexation policiesfor first-tier benefit levels as well as for public spending.

Chapter 3 focuses on how incomplete careers affect pension entitlements, focusing on delayedlabour market entry and career interruptions related to child-care, flexible and more precariouswork, and discontinued careers that are associated with earnings losses and lower pensions. Thus,with slightly more than a 1 per cent drop in old-age pension for every year without a job onaverage, pension systems can play a key role in offsetting potential losses in retirement income.But pension systems are not specifically designed to fully offset such contribution gaps, which varywidely across countries. This situation calls for a more comprehensive and integrated approach tohighly diverse life-cycle income profiles, involving a longer education cycle, wide skill-based paydifferentials, more complex family patterns, lower fertility rates, greater employment opportunitiesfor women, a longer (and rising?) average life expectancy, and the higher incidence ofdiscontinued careers with varying life-time earnings. In the absence of mechanisms to counteractthe impact of these trends on pension entitlements, the move towards DC pensions and the closerlinks between benefits and contributions are bound to affect income adequacy in retirement andheighten the old-age poverty risk. Pension credits can partly respond to this concern, but moreneeds to be done, including better interaction between public and private pension schemes andbetter work incentives. As well as contributing to ensuring longer and better working livesaccording to individuals’ preferences and choices, people will need to take more responsibility fortheir pensions, to which the design of pension systems – and labour market policy and legislation –

also need to be adapted.Chapter 4 examines how the changes in various OECD economic assumptions in this report affect

theoretical replacement rate estimations. Beyond the impact of mortality rates, these parameters relateto price inflation, real wages, rate of return on pension assets, and real-GDP growth. The replacementrates in DC schemes are generally more sensitive to changes in the values of the economicparameters of the model used as compared to either DB or flat-rate schemes. It notes thatreplacement rates increase, on average, 1.5 percentage points for the full-career average workeracross OECD countries. The majority of countries are affected by changes in the real-wage growthassumption.

Chapter 5 considers in detail five indicators in the design of national retirement systems in OECDcountries and other major economies. The first describes the taxonomy of the different kinds ofretirement income programmes worldwide, using this framework to describe pension systems in42 countries. The other four indicators deal with the parameters and rules of pension systems,starting with the second indicator that relates to basic, target and minimum pensions – their valuesand coverage. The third indicator looks at earnings-related pension systems – showing how benefitsare determined and the range of earnings covered. The fourth indicator highlights current retirementages by pension scheme for an individual entering the labour market at age 20 and retiring in 2014.The fifth indicator considers future normal and early retirement ages for an individual entering thelabour market at age 20 in 2014 and retiring in the future.



111

Chapter 6 looks at pension entitlements, using OECD cohort-based pension models, applying thesame methodology and assumptions to all the countries analysed to facilitate comparability of futureentitlement under existing parameters and rules (adopted up to June 2015). These comparisons arebased on the following indicators: (i) gross replacement rate in mandatory pension schemes (theratio of pensions to individual earnings); (ii) replacement rates for mandatory and voluntary pensionsschemes where these have broad coverage; (iii) the tax treatment of pensions and pensioners; (iv)net replacement rates, taking into account taxes and contributions; and (vi) pension wealth (in grossand net terms) – the life-time discounted value of the flow of retirement benefits, taking into accountthe retirement age, indexation of benefits, and life expectancy. As economic variables such aseconomic growth, rate of return on financial assets, real wage growth, the discount rate and inflationvary across countries, a single set of assumptions concerning these variables is used, so that crosscountry pension level differences reflect differences in pension systems’ policies only. It may beworth noting that, for workers with average earnings, the future gross replacement rate averages53 per cent for men and 52 per cent for women in the 34 OECD countries, obviously with substantialcross-country differences. Indeed, replacement rates vary between around 25–30 per cent of averageearnings for people starting work today in Mexico and the United Kingdom, compared to slightlyabove 90 per cent in the Netherlands.

Chapter 7 deals with population ageing, which is the main driving force behind recent pensionpolicies and reforms. It results from two demographic changes, namely, the decline in fertility ratesover the past 50 years and the increase in life expectancy. The latter is expected to continue, witha large variation between OECD countries in life expectancy at older age, and a gender gap inlongevity at age 65 that is expected to remain constant in relative terms, but to increase in absoluteterms (from 3.4 to 3.9 years on average). Given this trend, several governments have increasedtheir pension ages (or plan to do so), or introduced elements in their retirement income provision toautomatically adjust the level of pensions as people live longer. The indicators used are: (i) fertilityrates – the number of births and how this has evolved over the past 50 years; (ii) changes in lifeexpectancy at birth and at age 65; (iii) demographic dependency ratios – historical and projectedvalues 1950–2075; (iv) employment rates of older workers; (v) effective age of labour market exitversus pensionable age in 2014; and (vi) expected years in retirement from the time of averagelabour market exit by gender (in 2014).

Chapter 8 focuses on the incomes and poverty of older people – comparing these with thepopulation as a whole, while indicating the various income sources of older people (from publicly-provided benefits, earnings or private pensions, or other savings). It shows that incomes of olderpeople are on average lower than those of the population as a whole, even when account is takenof different household size. Those aged 65+ had incomes of 87 per cent of the total population’sincome, on average, in 2012, with big country variations (e.g. aged 65+ incomes were equal to orhigher than those of the total population at 100 per cent and 106 per cent in France andLuxembourg, respectively, closely followed by Greece, Israel, Italy, Mexico, Portugal and Spain withincomes above 95 per cent of the national average, but were down to just 67 per cent and60 per cent respectively in Australia and the Republic of Korea). People aged 66–75 have higherrelative incomes, on average, than those aged 75+, respectively 92 per cent and 80 per cent ofpopulation’s income. This is partly explained by cohort effects (e.g. due to growth of real earnings),which has led to higher earnings for each successive cohort of retirees, leading to higher pensionincome for each generation. Hence the importance of indexation for protecting pension income



112

over longer periods of time – notably for older women, who tend to have lower wages at work and longerlife expectancy compared to men. However, on average across the OECD, 12.6 per cent of individualsaged 65+ live in relative income poverty versus 11.4 per cent for the population as a whole. Yet, in18 out of 34 OECD countries, old-age income poverty is lower than for the population as a whole.

Chapter 9 analyses the finances of retirement-income systems, startingwith an overviewofmandatorypension contributions byworkers, then turns to public expenditure on pensions as a proportion ofGDPandof government budget, followed by private spending and total benefit spending on mandatory, quasi-mandatory and voluntary private schemes. The chapter concludes with long-term projections of pensionspending and, in particular, public expenditure on pensions between 2013–2015 and 2050. It looks atthe contribution side of pension systems to assess how much average workers contributed towardstheir pension in 2014 (tax-financed pension benefits were not included in this exercise). As differentpension components can be financed from different income sources, it is important, though sometimesdifficult, to map out the contributions that reflect the pension modelling. This differs from the totalcontributions from employers and employees, or via the social insurance or taxes that apply towage cost. Countries with higher pension contribution rates often have above-average pension benefits(e.g. Iceland and the Netherlands) or longer duration in retirement, owing to lower retirement ages (e.g.France and Italy). But higher contribution rates might lower overall employment and increase informality.It is worth noting that, across the OECD, public spending on cash old-age pensions and survivors’benefits increased 28 per cent faster than domestic output between 1990 and 2011, from an average of6.25 per cent of GDP to 7.95 per cent. Italy spent 15.85 per cent of GDP on public pensions in 2011, thelargest proportion across the OECD. Other high EU spenders include Austria, France, Greece andPortugal, at about 13–14 per cent of GDP, versus Germany, Poland and Slovenia at about 11 per centof total public expenditure. So, public pensions are often the largest single item of social expenditure,which accounts for 18 per cent of total government spending on average.

Chapter 10 looks at seven indicators to assess the situation of private pensions and publicpension reserve funds, more specifically: the proportion of the working-age population covered byprivate pensions (distinguishing between mandatory, quasi-mandatory, and voluntary schemes;and between occupational provision through an employer-provided or industry-wide scheme; andpersonal provision through a pension provider). The institutional structure of private pensions thatfollows distinguishes between pension funds, book reserves and insurance contracts – consideringseparately DB, DC and mixed or hybrid pension schemes. It also looks at private pensions andpublic pension reserves for 2013, the way these assets were invested and the investmentperformance of private pensions and public pension reserves in 2012–2013. It also includes DBfunding ratios for the same years.

To sum up, the report contains a wealth of information on cross-country pension policies andreforms, their achievements and shortcomings. It offers a useful tool for policy-makers faced withcomplex policy decisions to improve pension sustainability and adequacy in ageing societies andaddress the challenge of ensuring decent life in old age.

Hedva SarfatiIndependent consultant on labour market and welfare reform,Geneva, Switzerland



113

Acknowledgements

The Editorial Board thanks the following persons for their assistance in 2016:

Dragos Adascalitei, Nikola Altiparmakov, Assia Billig, Patricia Frericks, ElaineFultz, Aaron Grech, Igor Guardiancich, Riad Haddouz, Karl Hinrichs, DanielHogan, Stephen J. Kay, Charles Knox-Vydmanov, Mitchell Orenstein, FransPennings, Artiom Sici, Kari Tapiola, Johan Verstraeten.



115

CALL FOR PAPERS

Open call for papers

Triggered by a strategic realignment of the work programme of the InternationalSocial Security Association that gives greater priority to core questions of theadministration of social security schemes, the editorial team of InternationalSocial Security Review is launching an open call for high quality submissions ofinterest to an international readership.

The core area of interest is for critical analytic research output that investigateshow social security administration, supported by the innovative use of ICT, canstrengthen coverage and improve benefit adequacy in a sustainable manner.

With many national labour markets confronted by high levels of informality orwith increasing levels of precarity associated with a weakening of the traditionalemployer/employee relationship, papers are sought that examine options totransform social security institutional practices, financing mechanisms andbenefits delivery.

Authors are informed that there is no closing date for submissions.Submitted papers should be addressed to the Editor at: [email protected].



117

International Social Security

Review

VOLUME 70 | NUMBER 1JANUARY–MARCH 2017

The social security number: A small device underpinning big systems Paul-Anthelme Ad è le 3

Survivors' pensions and their impact on the Brazilian labour market Rogerio Nagamine Costanzi, Graziela Ansiliero and Julimar Da Silva Bichara 19

Worldview: An introduction to social security in Georgia Roddy McKinnon 49

The evolution of the Georgian social protection system since independence:An unclear future despite recent expansion Dimitri Gugushvili 51

The challenge of pension reform in Georgia: Non-contributory pensionsand elderly poverty Tamila Nutsubidze and Khatuna Nutsubidze 79

Book review 109

Acknowledgements 115

Call for papers 117

Information for subscribers: International Social Security Review is published four times a year in January, April, July and October. Subscription prices for 2017 are: Premium Institutional: £305 (UK), €388 (Europe), US$509 (The Americas), US$215 (Developing World), US$593 (Rest of World); Personal: £55 (UK), €80 (Europe, Euro zone), £55 (Europe, non-Euro zone), US$86 (The Americas), £55 (Rest of World). Prices are exclusive of tax. Asia-Pacifi c GST, Canadian GST/HST and European VAT will be applied at the appropriate rates. For more on current tax rates, please go to www.wileyonlinelibrary.com/tax-vat. The price includes online access to the current and all online back fi les to January 1st 2013, where available. For other pricing options, including access information and terms and conditions, please visit www.wileyonlinelibrary.com/access.

Delivery Terms and Legal Title. Where the subscription price includes print issues and delivery is to the recipient’s address, deli.very terms are Delivered at Place (DAP); the recipient is responsible for paying any import duty or taxes. Title to all issues transfers FOB our shipping point, freight prepaid. We will endeavour to fulfi l claims for missing or damaged copies within six months of publication, within our reasonable discretion and subject to availability.

Copyright and Copying: International Social Security Review © 2017 International Social Security Association. All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form or by any means without the prior permission in writing from the copyright holder. Autho-rization to photocopy items for internal and personal use is granted by the copyright holder for libraries and other users registered with their local Reproduction Rights Organisation (RRO), e.g. Copyright Clearance Center (CCC), 222 Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com), provided the appropriate fee is paid directly to the RRO. This consent does not extend to other kinds of copying such as copying for general distribution, for advertising and promotional purposes, for republication, for creating new collective works or for resale. Permissions for such reuse can be obtained using the RightsLink “Request Permissions” link on Wiley Online Library. Special requests should be addressed to: [email protected]

Back issues: Single issues from current and recent volumes are available at the current single issue price from [email protected]. Earlier issues may be obtained from Periodicals Service Company, 351 Fairview Avenue – Ste 300, Hudson, NY 12534, USA. Tel: +1 518 822-9300, Fax: +1 518 822-9305, Email: [email protected].

Journal Customer Services: For ordering information, claims and any enquiry concerning your journal subscription please go to www.wileycustomerhelp.com/ask or contact your nearest offi ce:Americas: Email: [email protected]; Tel: +1 781 388 8598 or 1 800 835 6770 (toll free in the USA & Canada).Europe, Middle East and Africa: Email: [email protected]; Tel: +44 (0) 1865 778315Asia Pacifi c: Email: [email protected]; Tel: +65 6511 8000Japan: For Japanese speaking support, Email: [email protected]; Visit our Online Customer Help available in 7 languages at www.wileycustomerhelp.com/ask

Abstracting and Indexing Services: ABI — INFORM (American Business Information); ASSIA (Applied Social Sciences Index & Abstracts); AgeLine; CSA Worldwide Political Science Abstracts; Environmental Sciences and Pollution Management; Health and Safety Science Abstracts; Insurance Periodicals Index; Medical Care Research and Review; PAIS International in Print (Public Affairs Information Service); Risk Abstracts; Russian Academy of Sciences Bibliographies; Social Services Abstracts; Social Insurance Research Network; Social Work Abstracts.

View this journal online at wileyonlinelibrary.com/journal/issrPrinted in Singapore by C.O.S. Printers Pte Ltd.

Production Editor: Niño Nestor Raynes (email: [email protected])Advertising: Joe Adams (email: [email protected])

Wiley’s Corporate Citizenship initiative seeks to address the environmental, social, economic, and ethical challenges faced in our business and which are important to our diverse stakeholder groups. Since launching the initiative, we have focused on sharing our content with those in need, enhancing community philanthropy, reducing our carbon impact, creating global guidelines and best practices for paper use, establishing a vendor code of ethics, and engaging our colleagues and other stakeholders in our efforts. Follow our progress at www.wiley.com/go/citizenship.

Wiley is a founding member of the UN-backed HINARI, AGORA, and OARE initiatives. They are now collectively known as Research4Life, making online scientifi c content available free or at nominal cost to researchers in developing countries. Please visit Wiley’s Content Access – Corporate Citizen-ship site: http://www.wiley.com/WileyCDA/Section/id-390082.html

Disclaimer: The Publisher, the International Social Security Association and Editors cannot be held responsible for errors or any consequences arising from the use of information contained in this journal; the views and opinions expressed do not necessarily refl ect those of the Publisher, the International Social Security Association and Editors, neither does the publication of advertisements constitute any endorsement by the Publisher, the International Social Security Association and Editors of the products advertised.

International Social Security Review accepts articles for Open Access publication. Please visit http://olabout.wiley.com/WileyCDA/Section/id-828081.html for further information about OnlineOpen.

INTERNATIONAL SOCIAL SECURITY REVIEW (Print ISSN: 0020-871X; Online ISSN: 1468-246X), is published four times a year in January, April, July and October. US mailing agent: Mercury Media Processing, 1850 Elizabeth Avenue, Suite #C, Rahway, NJ 07065 USA. Periodical postage paid at Rahway, NJ. Postmaster: Send all address changes to International Social Security Review, John Wiley & Sons Inc., C/O The Sheridan Press, PO Box 465, Hanover, PA 17331 USA.

The social security number: A small device underpinning ...

Documents

Transcript of The social security number: A small device underpinning ...