BookPDF Available

Rules of Relief; Institutions of Social Security, and Their Impact



This study is concerned with the collective consequences of social rules. The theoretical part takes as a starting point the body of ideas that has recently been developed in the 'new institutionalism' in the social sciences. The analyses focus on the nature and societal role of institutions, in particular where they relate to social security: the 'rules of relief' that are constructed by the community. Among other things, such institutions specify the benefits to which people may be entitled and the duties that are associated with the attribution of social security rights. The empirical part explores some of the effects of modern social security institutions. The central question is which collective outcomes are generated by social security regulations in 11 countries. The analyses first assess whether the institutions actually diverge so much that the national systems can in practice be said to represent three different types of social security: the liberal, the corporatist and the social-democratic model. Attention then turns to the impact of these three social security regimes. This investigation is performed on the basis of two indicators of collective output: do the regime types differ in the number of benefit recipients they generate? And is there a relationship between the empirical models of social security and the degree of poverty that occurs in the various countries? REVIEWS: “Rules of Relief is undoubtedly state of the art research in many ways… [The] book is a seminal contribution to the literature on institutional research and comparative welfare analysis. It contains a provocative analysis that will be useful for graduate students interested in social policy, institutional analysis and social security studies in general. Rules of Relief should also serve as a great reference library for researchers interested in understanding welfare state development in advanced industrialized countries.” − Michael Kpessa, European Journal of Social Security “The book shows an impressive knowledge and understanding, in a welfare perspective, of many elements and details of social security systems... It includes a central and important distinction between formal and informal social security and a discussion of the interaction between rules and behaviour, including compliance with certain sets of rules and decisions. This is perhaps one of the most neglected areas of social security studies.” − Bent Greve, Journal of Social Policy “Rules of Relief provides an interesting and a provocative demonstration that ‘institutions matter’ by outlining social security regulations in eleven developed countries. . . . The book provides a theoretical analysis of the origin, emergence and development of social security institutions as well as the empirical evidence of their impact corresponding to the three models. . . . [This book] provides meaningful insights to the collective significance of social security institutions and would be useful for researchers and students interested in social policy, understanding welfare state development in advances industrialized countries and making institutional analysis of social security issues.” − Mala Kapur Shankardass, Comparative Sociology
Rules of Relief
Rules of Relief
Institutions of social security, and their impact
J.C. Vrooman
The Netherlands Institute for Social Research | 
The Hag ue, September 2009
The Netherlands Institute for Social Research |  was established by Royal Decree of March 30, 1973 with the
following terms of reference:
a. to carry out research designed to produce a coherent picture of the state of social and cultural welfare in the Nether-
lands and likely developments in this area;
b. to contribute to the appropriate selection of policy objectives and to provide an assessment of the advantages and
disadvantages of the various means of achieving those ends;
c. to seek information on the way in which interdepartmental policy on social and cultural welfare is implemented with
a view to assessing its implementation.
The work of the Netherlands Institute for Social Research focuses especially on problems coming under the
responsibility of more than one Ministry. As Coordinating Minister for social and cultural welfare, the Minister for
Health, Welfare and Sport is responsible for the policies pursued by the Netherlands Institute for Social Research.
With regard to the main lines of such policies the Minister consults the Ministers of General Aairs; Justice; Interior
and Kingdom Relations; Education, Culture and Science; Finance; Housing, Spatial Planning and the Environment;
Economic Aairs; Agriculture, Nature and Food Quality; and Social Aairs and Employment
© J.C. Vrooman an d The Neth erlands Institut e for Soc ial Resea rch | , Th e Hague 20 09 - 2011
- publicat ion 2009-11
: Tex tceter a, The Hag ue
Figur es: Infor mation De sign Stud io, Amste rdam
Cover design: B ureau Stij lzorg, U trecht
Cover illust ration: Rel ief Bl ues (ca 1938), by O.L. Guglielm i (1906-1956). Temper a on berbo ard (61.1 by 76 .2 cm);
from t he colle ction of the Smith sonian A merican Art Mus eum, Wash ington D C,  .
With meticulo us realis m, Gugli elmi show s a famil y gather ed around a table i n a small Ne w York apar tment
durin g the Grea t Depres sion. A r elief wor ker lls o ut forms t o determ ine whet her they qualify for a welf are
bene t. Gugl ielmi por trays t he uneas e and des pair of th ose le u nemploye d, and th eir aem pts to sa ve face
and ma intain so me dignit y. The pa inter app lied for re lief hims elf duri ng the ea rly 1930s, bef ore he man aged to
obta in a meagr e stipen d through the Work s Progre ss Admini stratio n’s ‘Feder al Art P roject , a governme nt-
fund ed progr amme whic h at the t ime creat ed 5,00 0 jobs for artist s.
 978 90 377 0218 7 (pape rback)
 978 90 377 0431 0 (ha rdcover)
 740
Distr ibution in the  , Canad a and Mex ico: Transa ction Pu blisher s, New Br unswick ()
The Ne therlan ds Instit ute for So cial Res earch |  
Parna ssusple in 5
2511   The Hag ue
The Ne therlan ds
Tel.: +31 70 340 70 00
Fax.: +31 70 340 7044
Intr oduct ion
Inst itutio nal th eory 
. Institutions a nd the ‘new institutionalism’ 
.. Soc iological institutiona lism 
.. Rati onal choice institutionalism 
.. Histo rical institutionalism 
. North’s economic-historical ap proach 
.. Ins titutions and transac tion costs 
.. For mal and informal rules 
.. Enforce ment 
.. Ins titutional change and path dep endence 
. The evolveme nt and impact of institutions : a gural model 
. Institutions a s socially construct ed rules 
. Types of institution 
. Institutional h ierarchy 
.. Cons titutive and regulative ru les? 
.. Co re rights and derivati ve rights 
.. Hie rarchies of formal rules 
.. Hie rarchies of informal rules 
.. Co rrespondence betw een formal and informal rules 
. Actors , their relationships and their m otivations 
.. Ac tors 
.. Relat ionships between ac tors 
.. Moti vations of individual acto rs 
.. Moti vations of corporate ac tors 
. Rule-driven int eractions 
.. Inst itutions and interacti on 
.. Inte raction results an d the context of rule applic ation 
.. Rule a cquisition 
. The rule gene ration process: creation a nd development of instituti ons 
.. The in centive to regulate 
.. Und erlying causes of inst itutionalisation: the histo rical process 
.. Th e regulatory aspiratio ns of actors 
.. Rule -interaction and instit utionalisation 
. Conclusions 
Soc ial sec urity and t he inst itutio nal ap proach 
. Traditional denitions of s ocial security 
.. The na rrow approach 
.. The b road approach 
.. Cri ticism of the narrow and broad approac hes 
. Social secur ity in an institutional se nse 
. Institutions a nd actors in social sec urity 
.. Inform al social security rul es 
.. Infor mal social security s ystems 
.. For mal social security 
.. Infor mal elements in formal so cial security 
.. So cial security act ors 
. Undersoc ialised social securit y 
. Models of rule -driven social securit y interactions 
.. Familial so cial security 
.. Co mmunal social securit y 
.. Infor mal occupational social se curity 
.. For mal demographic regulati ons 
.. Une mployment and social assis tance benet regulatio ns 
.. Sick l eave and disability regulat ions 
. The results o f social security rul es 
.. The hi storical backgroun d to rule application 
.. Co nsequences for actor s 
.. Co llective results 
. Institutional c hange: from informal to formal sy stems 
. Conclusions 
Reg imes of socia l secu rity 
. Theoretic al traits of social sec urity regimes 
. An empirica l typology 
. The general ity of regime types 
. Conclusions 
Ben efit d epend ency 
. Regimes and b enet dependenc y: theoretical exp ectations 
. Measuring ben et dependency 
. The developm ent of benet dependenc y 
.. Relat ive volume 
.. Grow th rates 
.. Volume c omposition 
.. Summar y of the bivariate analy ses 
. A causal mo del for benet dependen cy growth 
. Country-s pecic models 
.. Impa ct of model variables 
.. Re gime impact 
. Conclusions 
Pove rty 
. The theoret ical poverty deb ate in political philosophy : some key elements 
.. Nature of t he decits: equalit y of what? 
.. Ab solute or relative decit s 
.. Est ablishing thresholds 
.. The n eed for compensation 
.. Con ditions for compensation 
. The meaning of pov erty 
.. Basic p rinciples 
.. A th eoretical denition of p overty 
.. Pove rty, inequality and so cial exclusion 
. Granting right s to the poor 
.. Comp ensatory right s 
.. Th e role of the government 
.. Co nditioning 
. Operationa l poverty lines 
.. Notio ns on poverty among th e population 
.. A t ypology of opera tional poverty line s 
.. Rel ative poverty lines 
.. Subj ective povert y lines 
.. Ob jective absolute pove rty lines 
. A generalise d budget approach 
.. Refe rence budgets for a single pe rson 
.. Init ial poverty lines 
.. Th e indexation method 
.. Ou tcomes of the generalise d budget approach in the Nethe rlands 
. The theoret ical relationship betwe en regime types and p overty 
. Empirical resul ts 
.. Count ry-specic norm amo unts for a single person 
.. Sen sitivity analysis of equi valence scales 
.. The ‘ three I’s of poverty’ a nd the regime types 
.. Ex plaining the poverty in cidence: multi-level model s 
. Conclusions 
The collec tive signif icanc e of so cial se curit y inst itutio ns 
. Institutions a nd social security 
. Regimes, b enet dependenc y and poverty 
. The impact of s ocial security regim es 
. Some implicat ions 
Ack nowled gement s 
Note s 
Lis t of ref erence s 
  
1 Introduction
Prior to the emergence of the modern welfare state, agencies taking care of the poor
sometimes drew up ‘rules of relief’. A Dutch wrien code dating from 1817 is exemplar y
in this respect. These ‘Regulations governing the provision of care to the home-dwell-
ing poor in the town of Den Bosch’ established the conditions for local poor relief. In
order to obtain support, residents had to have been affected by certain specific events.
The Regulations stated that a permanent need for municipal assistance could be due to
“advanced age and physical defect, which either in whole or in part makes the person
unsuitable for the performance of labour”. Temporary need might arise in the event of
“illness of short duration, cessation of work due to the season, women in times of con-
finement, and funeral expenses”.
However, not ever yone who became destitute due to such circumstances was entitled
to poor relief. The local officials caring for the poor (armverzorgers) could only nominate
people for support aer they had ensured that the other conditions set out in the guide-
lines had been met as well. Above all, they had to ascertain the “virtuous conduct of
those requesting help; though interestingly enough, the Regulations failed to specif y
precisely what criteria should be applied in making such a judgement. In addition, to be
eligible for help people had to have no ways of obtaining adequate means of subsistence
from their blood relatives, from their town of birth if this was somewhere other than Den
Bosch, or from other funds and subsidies. The ar mverzorger s also had to assess the degree
of “industriousness” displayed by applicants in partially meeting their needs themselves,
as well as whether their income could not be increased by “diligent labour”.
Following the preliminary investigation by the officials, the poor were “sampled”.
Twice a year the regents of the municipal poor relief board, the auditors, and the masters
of the ten local districts paid joint house calls to needy persons. Subsequently, the Coun-
cil of Regents decided which individuals and families would receive support, and how
much this would amount to each week in the coming summer or winter season. The arm-
verzorgers were responsible for the distribution of weekly poor relief; and they were only
allowed to increase the amounts set by the Council if unforeseen and exceptional costs
arose. In their contacts with the poor, however, these local officials had considerably
more latitude. They could restrict support as they saw fit, and largely at their own discre-
tion. Poor relief could be withheld for one or more weeks if the recipients “display con-
tinuous bad behaviour, or show carelessness towards their children, or on other grounds
which are deemed valid by the Lord Guardians of the Poor”.
As time passed the protection against the consequences of old age, illness, unemploy-
ment and similar risks came more and more to be regarded as a national responsibility.
The government was given a core task in safeguarding the living conditions of its citi-
zens. In order to achieve this, the favours granted by municipal poor houses and church
organisations to the deserving poor were gradually converted into the rights of the wel-
fare state, and enshrined in national law. De Swaan (1988: 218-257) discusses this histori-
1cal transition “from charity to social consciousness” extensively. In his view changing
social circumstances in particular the introduction of capitalist production methods,
the ongoing “civilising process”, the growth of the state apparatus, and innovations in
administrative techniques went hand-in-hand with changes in the configuration of
influential social groups (the working class, the pey bourgeoisie, employers’ organisa-
tions). The contesting elites went in search of new, stable ways to resolve the social issues
of their age, building on what had already been achieved.
De Swaan stresses the importance of the increasingly long chains of interdepend-
ence in this regard. Economic markets were no longer local or regional, but acquired
a national or even transnational character. On the social front, large-scale migratory
movements, the erosion of the traditional structure of social estates (nobility, clergy,
commoners), and the emergence of the nation-state meant that the mutual ties bet ween
individuals extended over greater geographical and social distances than previously. This
growth of interdependence implied that solutions for the problems social security aimed
to solve especially the economic, political and health threats originating from a large
group of urban poor – had to be sought at ever higher levels of abstraction. Initially the
local parishes and guilds bore the brunt of this collective effort; later the burden shied
to the municipal authorities and national mutualités, before finally ending up w ith the
modern welfare states, as yet the climax of the historical development.
Through this “collectivisation of care”, the rules of relief became more comprehensive
over the years. Not only did their number and complexity increase, but above all their
scope. As before, the government took responsibility for the alleviation of severe cases of
indigence; but it also began guaranteeing its citizens rights to a certain level of income,
medical assistance, employment counselling, social work, etc.
In essence, however, the main components of the rules have not changed. Just as in
the Den Bosch Regulations from 1817, they indicate which events entitle people to a cer-
tain benefit or provision, and how far those rights extend. And the contemporar y ‘rules
of relief’ also still contain clauses on job search requirements, on means testing, on the
administrative procedures to be followed, and on the sanctions that officials are empow-
ered to impose. It is modern social security rules such as these that form the subject of
this study.
To be more precise, the core focus is on the societal consequences of the institutions of
modern social securit y. Institutions are regarded here as socially constructed rules which
set out the rights and obligations of actors, and the associated conditions and sanctions.
These rules may be formalised in laws or government regulations, or laid down in con-
tracts between the different parties involved. They may also be more informal in nature
and usually remain unwrien: people’s mutual views on what constitute correct forms of
behaviour in particular circumstances.
On the one hand, institutions are a collective given. People cannot choose the his-
torically developed rules of the society into which they are born, and may find it difficult
to extricate themselves from their controlling force – especially when they perceive the
  
rules as self-evident. On the other hand, institutions are also a collective product. Rules
are made by people, and oen reflect the goals, interests, etc., of their compilers. Their
continued existence depends on social recognition and acceptance: institutions which
are not endorsed or observed do not achieve what they were intended to, and may eas-
ily become meaningless. Since they are socially constructed, institutions can in theor y
be revised, though in practice this may not be realised without a struggle. Institutional
change generally requires complex negotiations bet ween stakeholders, with the formal
legislative process in political democracies being the characteristic example. Actors are
moreover not always free to develop entirely different sets of rules. They oen reason
from the basis of existing institutions, which may appear to them as a natural order, the
abolition of which could entail high material or social costs.
Since institutions lay down the rights and obligations of actors, they influence the interac-
tions of people and organisations. This has consequences both for those directly involved
(their social standing, professional career, income, profit, etc.) and for the community
within which the rules apply (the relationships between social groups, crime rates, the
education level of the population, etc.). Modern social security institutions are of par-
ticular importance for two reasons. First, current ‘rules of relief’ are large in number and
control the behaviour of many actors: benefit claimants, contributors, benefits agencies
and business organisations. For that reason, they may have far-reaching consequences
for society. Social security institutions can exert a major influence on the economy, being
reflected in the nation’s wealth, the level of consumption and business investments. But
there may be other implications, too: the degree of poverty and inequality, the health
status of the population, the demographic profile, political relationships and the occur-
rence of social unrest can all be affected by the way in which the social securit y system is
Secondly, modern social securit y rules are very deliberate social constructs. They
are the outcome of an intensive and lengthy process of political decision-making which
is aimed at realising certain collective goals that are perceived as desirable. That makes
the field a suitable domain for investigating the nature of institutions and their social
Researc h quest ions
There are t wo types of question at stake in this study. The first concerns the theoretical
status of social security institutions, the second their empirical impact in modern socie-
ties. From a theoretical point of view it is essential to start by analysing in general terms
what institutions are, how they may arise and influence people’s behaviour, and what
consequences they may have for the actors involved and their community. The obvious
follow-up question concerns the applicability of such general notions to the domain of
social security.
Once these theoretical questions have been dealt with, it is important to investigate
what influence social security institutions actually have in practice. There are many ways
to assess this, both in terms of potential causes (various social security rules) and pos-
1sible effects (different types of outcome). The empirical analyses performed here home
in on the causal relationship between coherent systems of formal social securit y institu-
tionsor ‘regimes’and two macro-level outcome indicators: benefit dependency and
Chapter 2 looks at general theoretical issues, aiming to clarify the nature of institutions
on the basis of the following questions:
What does the notion ‘institution’ entail?
What kinds of institutions can be distinguished, and how do they relate to each
How do institutions give direction to social interaction, and which actors may be
involved in this?
What consequences can rule-driven interactions have?
Under what conditions do institutions come into being, and what causes them to
Chapter 3 explores the main topic of this study, social security, once again from a theo-
retical perspective. First the meaning of the concept in the scientific literature is sub-
jected to critical analysis, resulting in an institutional definition of social security. The
same questions are then asked as in the preceding chapter, but this time specifically for
social securit y rules. Aention focuses particularly on the ‘interaction structure’ of dif-
ferent types of social security schemes, and on the societal effects that social securit y can
theoretically bring about.
The next three chapters concentrate on the empirical significance of modern social secu-
rity regulations. The main question is consistently concerned with the collective results
social security rules actually generate. The emphasis is on formal institutions: social
security rules that are drawn up or ratified by governments. As De Swaan (1989: 11, 13)
obser ves, such rules are currently structured largely at national level, and shielded from
outside influences:
Welfare states are national states, which are concerned with the care only of their own citi-
zenr y. [...] Stat es erect borders betwe en their terr itory and that of other states, and welfar e
stat es seal off their domain of care just as securely from foreig n people. [...] The welf are stat e
is by nature exclusive and ant i-international.
In view of this national structuring of formal social security schemes, the obvious proce-
dure is to investigate their impact by means of a country comparison. The focus of inter-
est here will not be the significance of separate rules or regulations, such as the benefit
conditions in unemployment insurance or the level of early retirement pensions. Rather,
the objective is to answer a more generic question: what results do the diverse formal
national s ystems of social security achieve not in theor y (de jure) but in practice (de facto)?
Chapter 4 first explores the existence of such systems empirically. The leading ques-
tion is whether a quantitative analysis of a large number of formal social security institu-
  
tions reveals any country clusters that are based on fundamentally diverging principles.
If this is the case, we may speak of different ty pes of social securit y regimes. Such regime types
can be regarded as abstract models of institutional variety. Esping-Andersen’s (1990)
distinction between corporatist, social-democratic and liberal ‘worlds of welfare’ has
become famous in recent years. The empirical adequacy of the t ypology has been tested
previously by many authors, but oen not in an entirely satisfactory way. One major
problem is the limited number of formal institutions that are included in these analyses,
which probably implies that the various national social security systems have been repre-
sented only partially. The main aim of the chapter is to provide a more elaborate empiri-
cal test, by investigating whether a wide selection of formal institutional traits shows
any consistent clustering across 11 nations. Of these countries, the systems of Belgium,
France, and Germany are regarded as corporatist regimes in Esping-Andersen’s analysis.
Denmark, Sweden, and Norway are examples of his social-democratic type; and the ,
Australia, Canada and the United Kingdom theoretically represent the liberal model. The
Netherlands is also included in the empirical analysis, as an example of a social security
system which is difficult to classif y from a theoretical point of view.
In addition to this, several other interesting issues will also be examined in some
degree. Are regime types stable over time? Are they mainly confined to social securit y,
the labour market and the tax system, or are regimes more general, in the sense that
national institutions in the fields of health, education etc. show a similar paern of
countr y clusters? And finally: to what extent do the formal regime types coincide empiri-
cally with informal institutions, or the prevailing notions on social security, labour and
taxation among the population?
Aention then turns to whether social security regimes produce divergent collective out-
comes. Chapter 5 introduces a ver y direct output indicator: the volume of social secu rity ben-
efits in various countries. The degree of ‘benefit dependency’ is an interesting dependent
variable, as it is closely connected to the rights and duties aributed in social security
regulations, and because it may affect economic growth, labour market behaviour, pov-
erty, inequality, etc.
The key question is whether social security regimes differ in the number of benefit
recipients they generate, in line with theoretical expectations. This issue will be explored
for the same set of countries representing the various regime types (except Norway), on
the basis of comparative benefit dependency data covering a twent y-year period (1980-
1999). First, the theoretical relationship between regime types and the production of
benefits is discussed and cast in a number of hypotheses. These refer to three aspects of
benefit dependency: the relative volume (the share of the population receiving benefit);
the annual rate of growth in the number of benefits; and the composition of the total
benefit volume in terms of various social risks (old age, unemployment, disabilit y etc.).
The theoretical expectations are then submied to an empirical test. The first issue at
stake here is whether there is a direct link between the various regime types and the
three forms of benefit dependency. Aer completing these descriptive analyses, and hav-
ing discussed the implications in terms of the hypotheses, a stricter empirical test is per-
1formed. A multivariate model is developed, which makes it possible to assess empirically
whether regime types contribute to the production of benefits aer controlling for the
influence of other factors, such as demographic differences between countries.
Chapter 6 looks at the relationship bet ween regime t ypes and another outcome indica-
tor, the deg ree of pov erty across countries. Combating poverty is a major objective of most
social security systems; and the various regime types choose different strategies in order
to realise this goal. Does this imply that the degree of povert y brought about by the three
‘worlds’ (plus the Netherlands) differs as one would theoretically expect?
Since povert y is a more complicated construct than the number of benefit recipients,
the analysis starts with a demarcation of the phenomenon. How is poverty best conceived
of from a theoretical point of view? Once this has been ascertained, various operational
poverty lines are discussed and evaluated; and following this assessment a new criterion
is proposed, which will be used for an empirical comparison of poverty in the same coun-
tries as in the previous chapters. The aim is to test a number of specific hypotheses on the
relationship between social security regime types and poverty. Do the exponents of the
corporatist, social-democratic and liberal regime types selected here var y in the degree
of poverty they bring about, and do they do so in a way one would theoretically expect?
Once again, this issue is first pursued in a descriptive manner, by inspecting the bivari-
ate relationship bet ween regime types and various poverty indicators. Subsequently the
hypotheses are tested in a more rigorous way, by applying multi-level analysis to the data.
Using this method enables the unique contribution of the regime types to the degree of
poverty to be determined, controlling for the effects of other factors at the micro- and
Chapter 7 brings together the main findings, and discusses some of the results and their
implications for social science and policymaking. This over view aempts to elucidate the
principal motif of the study: the collective significance of social security institutions.
      
2 Institutional theory
This chapter discusses institutions from a general conceptual and theoretical perspective,
following the questions formulated in the Introduction. First a description is given of the
way in which the concept is interpreted in the scientific literature, with an emphasis on
the ‘new institutionalism’ in the social sciences (§2.1). Subsequently Douglass North’s
theor y is considered, which views institutions mainly in a historical economic perspec-
tive (§2.2). Building on these insights, §2.3 outlines a general figural model of institu-
tions and their social context. Its different elements are then treated in more detail in
separate sections.
The idea that institutions should be seen as socially constructed rules is elaborated
in §2.4. It will be argued that they encompass a social consensus on rights, duties, condi-
tions and potential sanctions. The next two sections briefly describe the various t ypes of
formal and informal institutions (§2.5), as well as how they theoretically relate to each
other (§2.6).
The actors for whom the social rules are intended, and their motivations and mutual
relationships, are the focus of §2.7. Subsequently, §2.8 is concerned with the way in
which institutions may influence the behavioural interchange of actors, and the results
to which this leads. A special form of such rule-driven interaction is also discussed in this
section, namely the way in which actors acquire rules.
Institutions are not unchangeable givens, but the product of the behaviour of actors
in certain historic circumstances. The theoretical mechanisms underlying this rule gen-
eration process are the focus of §2.9. The chapter ends with a number of conclusions.
2.1 Inst itution s and the ‘n ew inst itution alism’
In ever yday parlance the term ‘institution’ oen refers to agencies with a social purpose.
The Concise Oxford Dictionary defines the term inter alia as “a society or organization
founded especially for the promotion of science, education, etc”. In the social scientific
counter part to this, reference is oen made to entities which fulfil core social tasks: mar-
riage, family, the Church, voluntar y associations, political parties, and so on. Zijderveld
(2000: 33) gives an example based on anthropological functionalism, and also indicates
the limitations of this teleological approach:
There are basic biological and soc ial needs which are sat isfied th rough act ions. These action s
are permanently needed for the satisfaction of the needs, and will in time become schemat ic.
That is, they gradually grow into collecti ve habits and paer ns of behavior. There is, for exam-
ple, the biological need for sexual intercourse. The behav ior th at sat isfies this need, grows
into a reg ular paer n – the instit ution of marr iage.
This is, of course, a rather a-historic al and i nstrumental explanation of the or igin of instit u-
tions. The or iginal motive of this instit ution may have been the societal re gulation of s exual
inter course, but when the inst itution e xists as an objec tive and autonomous structure, it will
trigger and then regulate other, possibly quite different needs – such as, for inst ance, the nee d
to stabili ze emotions of love and affection, and the nee d to dispos e of a stable par enting facil-
ity. Solidly objectified and autonomous institut ions may even liberate indiv iduals f rom their
prim ary needs, set them f ree to design motives and aims which in turn may create or trigger
new needs. For instance, the goal of a formal dinner par ty is usually not the satisfaction of the
need for food. [...] It is, in a sense, a Leerform, an empty form which is filled with other mot ives
and aim s, such as net working , flirting, gos siping, forging polit ical compromises, and making
Mafia deals.
In the lemma that Eisenstadt (1968: 409) wrote for the Inte rnational Enc yclopaedia of the Soc ial
Sciences, he first discusses the standard functionalist interpretation of institutions:
Soci al inst itutions ar e usually conceived of as t he basic focuses of soc ial org anization, com-
mon to all s ocietie s and dealing w ith some of the basic social problems of ordered s ocial life.
Three basic as pects of institutions are emphasized. First, the paer ns of behavior wh ich are
regulated b y instit utions … deal w ith some perennial, basic problems of any societ y. Second,
inst itutions involve the regulat ion of behaviour of indi viduals in society according to some
definite, cont inuous, and organi zed paer ns. Finall y, the se paern s involve a definite norma-
tive ordering and re gulation; that is, regulation is upheld by norms and by sanctions which are
legitimized by t hese norms.
In this approach the institutional domain of family and kinship provides the solution for
the social tasks of reproduction and the initial socialisation of children. The institu-
tion of education is the answer to the social problems associated with the transformation
from children to adults and with the transfer of cultural heritage. The institutions of the
economy regulate the production, distribution and consumption of goods and services.
Political institutions control the use of force, maintain internal and external order on a
society’s borders, and take responsibility for the achievement of collective goals (viz., the
definition of those objectives, as well as the mobilisation and allocation of the necessary
resources for their achievement). Cultural institutions are concerned with the creation
and preser vation of religious, scientific and artistic products (‘artefacts’), and with their
dissemination. Finally, there is a separate institution focusing on stratification: the distri-
bution of positions, rewards and resources among individuals and social groups.
Eisenstadt then distances himself from this functionalist approach, however. In his
view institutions do not exist because they meet the needs of individuals or societies, or
because they reflect universal psychological or ecological tendencies. Such an approach
too easily assumes that needs are homogenous, rather than acknowledging that they can
differ or even conflict between groups, or between individuals and society as a whole. In
addition, lile aention is given to the socially optimal degree of the fulfilment of needs,
and to alternative institutional solutions which may be effective in similar historical cir-
cumstances. Finally, functionalist theories have difficulty in explaining institutional
change. All too oen it is assumed that the same conditions that lead to the creation of
certain institutions also ensure that they will continue to exist ad infinitum. This need not
be the case, however: changes in the nature of the needs, or in the historical and struc-
tural context, can cause institutions to become dysfunctional. The theoretical task is to
indicate under which circumstances they survive, change or vanish; but this question is
frequently not asked in functionalist approaches.
      
According to Eisenstadt, institutions are not given, constant, self-contained enti-
ties”; he prefers to speak of a dynamic process of institutionalisation. This is not an expres-
sion of abstract social functionality, but refers to the sustained regulation of exchange
between actors. Institutionalisation can be seen as (Eisenstadt, 1968: 414)
A process of continuous crystalli zation of different ty pes of norms, organizations, and f rame-
works which regulate the pro cess of exchange of different commodities.
The exchange relates to different goods in the various institutional domains: in the
economy, for example, it refers to the strengthening or loss of market positions; in the
political field it has to do with the power that people acquire, the support they generate
and the compromises they are able to achieve. Institutionalisation of social exchange is
expressed among other things in legislation, communications systems, the administra-
tive organisation, and the regulation of economic and political markets.
In Eisenstadt’s vision institutional change is not a blind, unfocused process. It is instead
one which constantly builds on existing institutions (Eisenstadt, 1968: 415):
The concrete organiz ational structures in the prece ding situation [...] create the conditions
for their ow n change.
Exchange theory states that ‘institutional entrepreneurs’ play a key role in the process
of institutionalisation. These are people with a particular ability to articulate new goals
and norms, to set up new organisations and to mobilise the resources needed to ensure
their functioning. Their institutionalising capability is related to the control they have
over key positions and resources; however, having a high degree of control does not auto-
matically mean that they are influential in shaping institutions. Of crucial importance is
whether they are sensitive to societal needs, and that they are able to ensure that their
own solutions to those needs prevail. Institutional entrepreneurs aempt to achieve this
through a mix of coercion, manipulation and persuasion. They are not entirely free in the
choice of their goals and behaviours; the interests, the possibilities offered by their social
position and those of their rivals (power, money, contacts, communication channels),
their own views and those of their supporters, etc., limit the nature of the changes they
can seek to bring about. However, their course of action is not entirely predetermined:
entrepreneurs have some latitude, which in this approach is considered to be the most
important driver of institutional change.
In Eisenstadt’s definition, institutions comprise both the social rules for exchange,
and the organisational configuration which effectuate these. From an analytical point
of view it seems useful to separate these t wo aspects more explicitly. There are societal
rules, and the extent to which they effectively regulate human behaviour depends among
other things on collective organisational forms. However, these are two different things,
which do not necessarily lie on the same line. Rigorous norms may be accompanied by a
low level of collective organisation: e.g., the behaviour of scientists is governed by fairly
strict professional rules, which are supported by a loose (and partly anonymous) network
of peers throughout the world. It would be hard to maintain the argument that this weak
organisation implies institutions are nonexistent in such a case.
Throughout this study the concept therefore does not refer to key organisational
forms or to the deployment of social resources, but exclusively to the applicable social
rules. This fits in with a second everyday meaning of the term: “an established law, cus-
tom, or practice(Conc ise O xford Dictionar y). It also aligns with the interpretation of the
classical sociologist Durkheim, who saw institutions as manifestations of ‘social facts’,
or collective ways of acting, thinking and feeling. According to him, this supra-individ-
ual, objectified reality is the quintessential object of study for sociology (Durkheim, 1901:
On peut en effet , s ans dénatur er le sens de ce e expression, appeler institution, toutes les
croy ances et tous les modes de conduite institués par la collect ivité; la soc iologie peut alors
être défin ie: la science de s instit utions, de leur genèse et de leur fonct ionnement.
In Durkheim’s view, institutions offer a historically rooted pre-structuring of social real-
ity and are therefore a means of avoiding anomia (lack of norms, social chaos). Starting
from a different theoretical perspective, G.H. Mead (1934: 167, 211) arrives at a similar defi-
nition of institutions:
What we mean by [an instit utional form] i s that the whole communit y act s tow ard the indi-
vidual in an ident ical way … An inst itution is, aer all, not hing but an organi zation of ait udes
which we all carry in us, the organ ized aitudes of the others that control and deter mine
A definition of institutions in terms of collectively rooted convictions which influence
the behaviour of actors returns in a recent theoretical approach known as ‘new institu-
tionalism’. Hall and Taylor (1996) distinguish three variants of this: sociological, rational
choice and historical institutionalism.
2.1.1 Soc iologic al inst itution alism
Adherents of sociological institutionalism defy the functionalist approach, in which the
normative force of rules provides for social order. By contrast, the cognit ive significance
of institutions is emphasised. In line with social constructivism, of which the work of
Berger & Luckman (1966) is a well-known exponent, sociological institutionalism pos-
its that institutions do not consist exclusively of rules, procedures and norms. Above all
they offer symbol systems, cognitive interpretation frames, and moral templates which
guide behaviour: “Institutions influence behaviour not simply by specifying what one
should do, but also by specifying what one can imagine oneself doing in a given context”
(Hall & Taylor, 1996: 948). According to these authors, this has a number of implications.
In the first place it blurs the distinction bet ween bureaucratic rationalit y and culture’.
Following on from Weber, organisational sociologists for a long time argued that the
central role accorded to formal, rational rules and procedures in government agencies
and companies is an efficient adaptation to the task s that they fulfil in modern societies.
      
In this line of reasoning, culture – in the sense of shared beliefs or values was oen
placed outside the arena of formal organisations. A Wertrat ionalität (‘value rationality’)
was only considered important to explain the behaviour of voluntar y associations, ideo-
logical movements and religious sects; but the actions of modern, formal organisations
were driven in principle by Zweckrationalität (‘ends and means rationality’). Sociological
institutionalists counter this with the argument that all institutions have both formal
and informal aspects (Hall & Taylor, 1996: 946-947):
Many of the institut ional form s and procedures used by mo dern organiz ations were not
adopted simply because they were most efficient for the tasks at hand, in line w ith some tran-
scendent ‘ration ality’. Instead [... many of thes e] should be seen as c ulturall y-speci fic prac-
tice s, akin to t he myths and cer emonies devis ed by many so cieties, and as similate d into orga-
nizations, not ne cessar ily to enhanc e their formal means-ends efficienc y, but as a result of the
kind of process es as sociated with the t ransmission of cult ural prac tices more gener ally [...]
Even the mo st seemingly bureaucratic of practices have t o be explained in cultural t erms.
Secondly, in contrast to the rational choice institutionalist view (see below), in this
approach the individual is not an autonomous homo ec onomicus. Sociological institution-
alists have a dialectic perspective on the relationship between man and society. In the
words of Berger & Luckman (1966: 79):
Society is a human product. Societ y is an object ive realit y. Man is a soc ial product [...] An anal-
ysis of the soc ial world that leaves out any one of these three moment s will be distort ive.
Thus, cognitive frames of interpretation are created by people, but eventually acquire
universal applicabilit y. The transfer of these ‘ways of seeing’ in the socialisation process
ensures that new generations are familiar with the rules. And when those generations
apply the institutions, they confirm to both the social prescript itself and their member-
ship of social entities. Hall & Taylor (1996: 948) therefore designate the sociological- insti-
tutionalist perception of the relationship bet ween institutions and individual behaviour
as interactive and mutually constitutive.
Thirdly, according to this view the process of creating and changing institutions cannot
be explained entirely by the efficiency with which these achieve the instrumental goals
of actors. Rather, institutional change is oen induced by the appropriateness of certain
types of rules (Hall & Taylor, 1996: 949):
Organiz ations oen adopt a new institution al practice, not because it ad vances the means-
end efficienc y of the organiz ation, but bec ause it enhances the social legitimacy of the organi-
zation or its participants. In other words, organiz ations embrace specific inst itutional for ms
or practices becau se the laer are widely valued within a broader cultur al env ironment. In
some cases, these practice s may actually be dysfunc tional w ith regard to achieving the orga-
nization’s formal goals [...] This picture [may be captur ed] by de scribing it as a ‘logic of social
appropr iateness’ in contr ast to a ‘logic of in strumentality .
Finally, the legitimacy of institutional arrangements is not self-evident. The fact that
some institutions are regarded as socially appropriate while others are not has to do with
the organisation of cultural authorit y. Sociological institutionalism recognises different
sources of authority. The state and the political process can lend public authority to cer-
tain institutional arrangements through legislation or official recognition. In addition,
the growing number of experts (technicians, lawyers, physicians, economists, sociolo-
gists) can use their professional authority as a basis for imposing standards on their peers
and for influencing public policy relating to their field. Furthermore, institutions may
be the result of an interactive decision-making process by actors in certain influential
networks. Through peer group discussions they can reach a consensus on what are con-
sidered legitimate interpretations and solutions to social problems. Such an ‘interactive
legitimacycan occur, for example, in international organisations which issue more or
less mandator y recommendations as to which rules countries should follow in order to
achieve a balanced government budget, economic growth, educational reform, techno-
logical innovation, etc. (Hall & Taylor, 1996: 949-950).
2.1.2 Rational choice ins titutio nalism
The second variant, rational choice institutionalism, posits that individuals behave
instrumentally and strategically. They determine their choices independently on the
basis of their own preferences or tastes. These preferences are given, and individuals try
to achieve them to the maximum through strategic behaviour based on a meticulous
consideration of the possible costs and benefits of various alternatives. In this approach,
institutions are rules which provide confidence in people’s current and expected behav-
iour; in doing so, they lower the risks of commercial and social exchange. Institutions
provide information about what normal behaviour is, limit the number of choice options,
ratify contracts and agreements, impose sanctions when people defect, etc. Rules con-
tinue to exist as long as a substantial group of people believe they would be worse off
not conforming with the prescribed behavioural paern than by conforming to it. If this
social support dwindles, institutions will change if a more efficient alternative is avail-
able (see Hall & Taylor, 1996: 942-946; Brinton & Nee, 1998).
Rational choice theory has been widely used in studies on the problems associated with
collective action, such as political decision-making. Hardin’s Tragedy o f the commons (1968)
is the exemplary portrayal of the ‘social dilemma’. In this article he referred to the predic-
aments that arise in farming villages with common grazing land when the laer becomes
scarce. For each individual farmer it ser ves his self-interest to place as many animals as
possible on the collective farmland; but if all farmers do this, the commons will eventu-
ally perish due to over-grazing. This is the essence of a social dilemma: behaviour that is
rational for the individual has negative consequences for the community as a whole, and
is therefore irrational from the point of view of the collective interest. A literary example
of this mechanism can be found in Ian McEwan’s novel Enduring Love.
Social dilemmas are difficult to avoid when it comes to collective goods to which every-
one has access, and from which many people derive utilit y. The conceivable solutions
oen bring their own disadvantages. In a hierarchical approach the desired behaviour is
      
imposed by a strong central organisation (usually the national government). However,
this tends to be at odds with the individual freedom of action. In democratic societies, a
hierarchical solution is therefore only viable when the undesirable behaviour has conse-
quences for other parties which are generally deemed unacceptable.
Contractual solutions seek to avoid people acting as free-riders through agreements
between the direct stakeholders, with sanctions being imposed for infringement of the
contractual stipulations and compliance rewarded. However, this becomes a complex mat-
ter if many actors are involved, if the consequences of the behaviour only become appar-
ent aer a long period, or if the individual costs and benefits are difficult to establish.
If a solution is sought for social dilemmas via the market, an aempt is made to change
the costs/benefits ratio in such a way that it becomes rational for a critical mass of people
to focus on the common interest. Market-based solutions may relate to the immediate
costs and benefits of people’s actions (price interventions, prevention of cartel-forma-
tion), or to the transparency and accessibility of information (e.g. by prescribing that
companies must produce an annual report, and stipulating what they have to make pub-
lic in it), and the required control mechanisms (e.g. external approval by independent
auditors). However, the pricing of collective goods can meet with resistance, as for exam-
ple with the introduction of road tolls during peak hours, which are regarded by many
as a selective infringement of the right to mobility. Determining the optimum incentive
structure in such a case is also oen difficult, and introducing the regulation will involve
costs (e.g. for the technical infrastructure needed to enable tolls to be levied).
Finally, solutions sought via the community try to make the general interest prevail
through stimulating shared standards of conduct, promoting awareness of the disagree-
able social outcomes of defection, more intensive social control, and establishing overt
moral commitment. Examples include the promulgation of an official ‘catalogue’ of val-
ues and norms by the authorities; information campaigns targeted at the prevention of
the collective bad; the visible naming and shaming of defectors; and making people take
a public vow to behave as desired (no smoking, drugs or alcohol; abstaining from sexual
intercourse before marriage; respecting the environment). The main problem here is that
it tends to be difficult to impose shared standards, awareness, sanctioning and commit-
ment ‘from above’: ultimately, the members of the community themselves must support
such solutions and see them as self-evident. Where communal ties are weak or on the
wane, it is not easy for third parties (such as policy-makers) to assure compliance through
the community.
Schuyt (see Vrooman, 1999) stresses that the rational choice approach is aractive
because of its simple structure, which offers an explanation for much of the behaviour
of individual and collective actors. Many variants of the theoretical principles have been
studied in empirical research (e.g. single-person and multiple-person dilemmas, zero
and non-zero-sum games, situations with complete and incomplete information on the
implications of the choice process), and the theory has therefore been validated in all
manner ofelds. The theoretical model is applied in sociology to explain relationships
between individuals (the selection of spouses, friendship relations), between organisa-
tions (cooperation and competition of firms), between individuals and organisations
(hiring and firing of employees), between individuals and society (collective impoverish-
ment due to individual rational behaviour, as posited by Hardin), and between collectivi-
ties (e.g. tax policy competition by states).
However, Schuyt also points to a number of methodological and theoretical objec-
tions, which make it doubtful whether rational choice theory can be regarded as the
“quantum leap of the social sciences”. In the first place the notion of ‘rationality’ is oen
interpreted ver y broadly in this approach, so that for example altruistic behaviour (giv-
ing money to a beggar) becomes explainable as a rational choice (because of the moral
satisfaction derived from it). The theory then tends towards tautology, with all behaviour
being rational in the final instance. In Schuyt’s view, it is more common that behaviour
is determined by a mix of rational choice and other motives, such as trust. He also ques-
tions whether the opposition between the individual and the collectivit y is not over-
emphasised in rational choice theor y. For example, the right to Dutch citizenship does
have individual bearers, but would be hard to imagine without a shared collective his-
torical background; in reality, individual and society are oen co-constitutive. Finally, he
comments that in a rationalising societ y, rational choice theory may become something
of a self-fulfilling prophecy. The greater the extent to which actions are rationally driven,
the more types of behaviour can be explained by rational choice theor y.
In the economic sciences a variant of rational choice institutionalism became in vogue
in the 1990s, namely the ‘new institutional economics’ (), or the economics of trans-
action costs (cf. Williamson, 1985, 1998; Williamson & Winter, 1991). This harks back
to the classic work by Coase (1937, 1960), from which the proponents of the derive
their basic theoretical principle: when it is costly to transa ct, institut ions ma er. According to
this approach, institutions are more or less efficient solutions for economic coordina-
tion problems, since they lower transaction costs. Unlike neo-classical economic the-
ory, there is no assumption of actors behaving in a fully rational way, but rather of their
‘bounded rationalit y’, to quote Simon (1986). People do not necessarily wish to maximise
their utility; and in everyday situations their opportunities for rational action are oen
limited. In certain conditions actors may sele for ‘satisficing’ instead of maximising
utility: they try to achieve a level which they consider sufficient. Aiming for the satisfac-
tory theoretically prevails when choices relate to aspects that are less central to the actor.
For instance, a person may wish to maximise his income or happiness, but in purchasing
specific consumer goods (such as washing-powder) many different brands probably will
be good enough.
Perhaps even more important, the bounded rationality view stresses that people
are not always able to make rational choices. Actual behavioural choices are oen based
on a simplified view of reality. Actors tend to take only a limited number of factors into
account, which they think are most relevant or crucial to them; and in the process, they
are prone to make misjudgements. In some cases they may not be able to acquire all the
information that is relevant for making a rational choice; or they may not want to seek
this, since it requires too much time or money. In other instances, information may be
      
so abundant that people cannot process it adequately. Moreover, people are not always
able to determine the consequences of their behaviour accurately in advance, and their
preferences and interests may not be stable or clear. Andnally, their decisions may be
influenced by situational and personal factors (the degree of shade in the room, their
serotonin level, their predisposition to calculate, their abilities to do so).
However, in the bounded rationality perspective actors intend to make rational
choices, in the sense that they opt for actions which they regard as effective means to
accomplish their ends (interests or preferences). Schmidtz (1995: 12-13) points out that
people may act rational even if they ‘objectively’ make the wrong choices. What maers is
that an actor is aware of his ends; that his behaviour is directed towards these goals; and
that he has good reason to believe his particular line of action is effective enough.
2.1.3 Histori cal ins titutionalism
In the third approach institutions are fairly strongly allied to formal organisations, and
their creation in a given socio-historical context is emphasised. As Hall & Taylor (1996:
938) point out, institutions are then regarded as
Form al or infor mal proce dures, routines, norms and conventions embe dded in the org aniza-
tional str ucture. [...] They can r ange f rom the rules of a constit utional order or t he standard
operating proce dures of a bureaucracy to the convention s gover ning trade union behaviour
or bank-firm relat ions.
Historical institutionalists generally share their view of the relationship between people
and institutions with one of the other two schools mentioned above, sociological and
rational choice institutionalism (Hall & Taylor, 1996: 939-940). In some other respects,
however, their approach is a unique one. It is centred in the first place around the histori-
cal restraints of social evolution, as expressed in the notion of path depen dence. Historical
institutionalists distance themselves from the traditional view that similar developments
– e.g. technological innovations – will lead to the same results ever ywhere. On the con-
trary, they argue that the existing historical context operates in an intermediating way:
whether or not a given innovation is implemented, and the extent to which it catches
on, depends greatly on the characteristics of the society concerned. Institutions play an
important intervening role here: if the formal and informal rules are at variance with the
innovations, the laer will not be implemented as a maer of course.
A textbook example of technological path dependence is the sur vival of the 
standard for the keyboard layout in typing and word processing (David, 1985). In the
1860s, the first typewriters had the disadvantage that the impression made by the type
bars on the paper was not immediately visible. If one of the keys got stuck, this caused
the same leer to be typed over and over again, something which only became apparent
when the carriage was lied. The inventor of , Richard Sholes, tried to minimise
the number of collisions between the type bars by placing widely used leer combina-
tions in English a long way apart. At the time this was a technological improvement,
which was patented in 1868 and included in Remington’s first Type-Writer’ in 1872.
Already in the 1890s an alternative for was available: Blickensderfer’s ‘scientific
keyboard’, which used the -arrangement, and enabled 70% of English words
to be typed using the keys on the boom row of the keyboard. According to David 
nevertheless prevailed, mainly because users had already invested in the method: during
the period that the market for typewriters was growing rapidly, touch-typing on 
keyboards had become the standard. The costs of switching to a different system would
have been relatively high for users; while for typewriter manufacturers it required a fairly
small investment to incorporate   keyboards into their existing models (by solder-
ing the type to different bars and changing the keys on the levers). The continued exist-
ence of this ‘institution’ can be seen as an example of path dependence: every new key-
board layout first had to demonstrate that the advantages of its introduction out weighed
the drawbacks of abolishing the  standard (breaking through existing habits, the
time and costs of retraining users). Later superior alternatives like the Dvorak or Velotype
keyboards supposedly did not succeed due to path dependence, which leads David (1985:
336) to conclude that
Compet ition [...] drove the indust ry premat urely into standardization on the wrong s ystem
where decentralized decision making subsequently has sufficed to hold it .
Of course, path dependence becomes more complicated when applied to the analysis of
social processes. In historical institutionalism the theory has been developed mainly in
comparative studies of government policy and collective decision-making. Countries
may respond entirely differently to similar policy problems because the existing institu-
tions make some solutions appear more obvious than others. There may be several rea-
sons for this. The prevailing legislation and related informal rules can promote change,
but can also hinder it. For example, whereas unprofitable farms go bankrupt in free
market countries, they may be kept alive elsewhere via state subsidies, a practice that
is oen legitimised by referring to the national interest of an independent food supply,
or by pointing to the need to protect farm products that are considered unique. Similar
developments may also lead to different results if the shared views of policymakers (the
‘policy culture’) or the configurations of interests (e.g. strong or weak trade unions) var y
among countries. A factor of prime importance is that countries have invested in their
current institutions. This means that changing the formal rules can entail higher costs in
some countries than in others. Moreover, if the informal views in a country oppose such
a change of the formal institutions, it may prove difficult to implement it.
Hall & Taylor (1996: 941-942) conclude that the notion of path dependence implies
that institutions are not always functional:
Historic al institutionalists stress t he unint ended consequences and inefficiencies generated
by ex isting inst itutions, in contrast to images of institution s as more pur posive and efficient.
A second characteristic of historical institutionalism is the aention for the dis tribut ion
of power a nd the exist ence of c onflicts over scarce goods betwee n riv al groups. Institutions are not
neutral rules (Hall & Taylor, 1996: 941), but
      
[…] give some groups or interest s disproportionate access to the decis ion-making process;
and, rather t han emphasize the degree to which an outcome mak es ever yone be er off, [his-
torical inst itutionalist s] tend to s tress how some groups lose while ot hers win.
In the third place, this school of thought sees institutions as being not the only cause of
social change. Although the formal and informal rules determine the direction and scope
of changes in societ y, they do so in combination with other factors, such as demographic
and socio-economic trends, scientific and technological developments, etc. As a result,
society is “more complex than the world of tastes and institutions oen postulated by
rational choice institutionalists” (Hall & Taylor, 1996: 942).
2.2 Nor th’s econ omic-historic al approach
In his theoretical analysis Inst itution s, instit utional chang e a nd e conomic per formance (1990),
Douglass North combined several of the above insights, though he integrated the eco-
nomic and historical perspectives beer than the sociological aspects. North (1990: 3)
defines institutions as
the rules of the game in a s ociety, or, more formally, humanly dev ised cons traint s t hat shape
human interac tion.
Institutions may be formal (laws, regulations) or informal (social norms, customs,
behavioural codes). Infringing such rules incurs sanctions; this implies that the price of
establishing defections and the severit y of the punishment is an important part of the
functioning of institutions.
Institutions have to be conceptually separated from organisations. North considers
the former as “the rules of the game”, while the laer are among the players. There is a
dialectic connection between organisations and institutions. On the one hand the exist-
ing rules help determine which organisations are formed and how they develop; organi-
sational forms are a response to the incentive structure offered by institutions. On the
other hand, organisations also influence the development of institutions: they seek to
change the opportunit y structure in such a way that they derive more benefit from it.
2.2.1 Institut ions an d trans action co sts
Institutions regulate human behaviour, and North assumes that actors behave on the
basis of the bounded rationalit y perspective described earlier. He believes that transac-
tion costs play a key role in the regulation of behaviour. These expenses are the sum of
the costs of:
a) defining and protecting propert y rights, and monitoring and enforcing agreements;
b) measuring the valued aributes of goods and services.
Economists have traditionally focused mainly not on transaction expenses, but on trans-
formation costs: the investment of land, labour and capital that is needed to change the
physical characteristics of a good in such a way as to create added value. North, however,
points out that transaction costs in modern economies are considerable – for the United
States he estimates that they account for 45% of  and therefore have a substantial
influence on the economic process as well. And because institutions provide a framework
for economic exchange, they are important determinants of the level of transaction costs.
First and foremost, institutions regulate the way in which ownership of a good is
acquired, how it may be used or how its use may be denied to others, and how the con-
tract can be observed. For example, the propert y rights to a dwelling are generally highly
regulated in modern societies. There are formal rules dictating the content of the sales
contract, the entr y in the land register, agreements with the mortgage provider, the tax-
able status of home ownership and the tax deductibility of housing and maintenance
costs, the inheritance laws, the possibilities to evict and prosecute squaers, etc. These
rules are oen reinforced by the self-evident mutual expectations of buyers and sellers
on what is a reasonable margin between asking price and bid, the role of estate agents,
etc. Institutionalisation needs not be that high, however. In a societ y where tracts of
ter ra nulliu s are available, it may be that property rights are created by simply fencing off
or inhabiting such areas and defending that land against any competitors. But in both
cases establishing the property right involves transaction costs: the fees for the notary,
the mortgage provider and the tax office in the highly regulated society, and the costs of
defending one’s home in the other variant.
In addition, ever y economic exchange involves meas urement cos ts. The value of an
exchange commodity is the sum of the characteristics which are involved in the good
or the service for both parties. If someone buys a house, they purchase a number of
aributes: a certain number of square metres, a particular building style, the solidity of
the construction, the aractiveness of the neighbourhood, the distance from central
amenities, etc. To determine whether the desired characteristics are included in the sale
and whether the asking price is in line with the market, it is necessary to make inquir-
ies: comparing houses and neighbourhoods, carr ying out architectural research, etc.
This takes time and money, and in complex exchange it is oen not feasible to obtain all
relevant information. Moreover, information asymmetries oen occur in practice: the
vendor of the house generally knows more than the buyer, and either one may have an
interest in hiding or revealing certain traits.
The costs of establishing property rights and measuring valued aributes may vary
depending on how efficient the rules are. In general, total transaction costs increase as
the type of exchange and the socio-economic seing become more complex; and as the
ability of the actors to understand the context, and to enforce compliance with the rules,
2.2.2 Form al and infor mal rul es
North draws a distinction between formal and informal institutions. He suggests that the
difference is a gradual one: it may be regarded as a continuum of institutionalisation, with
customs, traditions and taboos at one extreme, and a wrien constitution at the other.
      
Formal rules are created in order to make exchange (in a political or economic sense)
possible, and are based on the initial negotiating power of decision-making parties. For-
mal rules need not necessarily be efficient: in principle they are designed more to foster
the interests of those in power than to optimise the general interest. North identifies three
types of formal institutions: political and judicial rules, economic rules and contracts.
1. Politic al and judicial rules. These reflect the hierarchy of the various policy issues, as
well as the key points of the decision-making procedure and the method of agenda for-
mation. Political and judicial rules have a certain primacy because they shape the play-
ing field for economic rules and contracts. This primacy is incomplete, however, because
there is also a feedback mechanism in which actors aempt to influence the political and
judicial rules.
In the simple case there is one dominant ruler, who in exchange for tax revenues dis-
penses justice, provides safet y (or at least prevents chaos) and protects property rights. As
different factions have different opportunit y costs and negotiating power, they enter into
individual contracts with the dominant ruler.
In a slightly more complicated model there is a representative body which acts in
behalf of the interests of a limited number of groups (such as the Spanish Cortes, origi-
nally a parliament of the three estates). This enables the ruler to generate more tax rev-
enues in exchange for the granting of privileges to the interest groups and their agents.
This creates a hierarchical structure and an extensive bureaucracy.
In modern representative democracies there is a multitude of interest groups and a
much more complex institutional structure designed to foster the exchanges between
the various groups. There is also not just a single legislator, but a legislative assembly. All
representatives (s) have to satisfy their own constituencies, each of which has its own
characteristics. This cannot be achieved through a simple exchange of votes; oen it is
a maer of making prior agreements on voting cooperation, or of granting each other
influence on certain closely defined topics and on the agenda (e.g. the chairmanship of
2. Economic rules. These mainly establish propert y rights. In the simplest model prop-
erty rights are a function of changes in economic costs and revenues: they arise when
changes in relative prices and/or scarcit y are such that they counterbalance the costs
of granting or enforcing the rights. There are however also many inefficient rights, the
monitoring of which is not economically viable. North explains this on the basis of the
inefficiency of the political markets: rule makers do not wish to upset their influential
supporters, or it may be that the costs of establishing and collecting levies are so high
that a less efficient allocation of rights leads to higher revenues. An efficient distribution
of property rights arises where the political transaction costs are low and the political
actors have adequate subjective perceptions. In practice this is oen not achieved.
3. Contract s. These contain the conditions governing specific exchange agreements. In
traditional economic seings contracts oen referred to the trading of a single good at
a single moment. Modern contracts cover several goods and extend over a longer period.
They are by definition incomplete and therefore stipulate which maers will be decided
through arbitration by a third part y or the courts.
According to North informal institutions, such as codes of conduct, norms and conventions,
are oen embedded in the formal institutions. In many respects, however, they are more
important, as formal rules contain few instructions for everyday behaviour. Their influence
is evident from the fact that the same formal rules can lead to different results in a different
social and historical context, and from the persistence of informal rules long aer a radi-
cal amendment of the formal institutions and their related organisational forms (e.g. the
Japanese culture aer the  administration following World War ; the Jewish and Kurdish
culture in the Diaspora, the Orthodox Russian Church aer 70 years of Communism).
North (1990: 37) broadly equates informal institutions to culture:
Infor mal constraints [...] are part of the heritage t hat we cal l culture, [... which] can be defined
as the tr ansmission from one generation to the ne xt, via teaching and imitation, of knowledge,
value s, and other fac tors that influence b ehavior.
A sociologist would probably be more inclined to refer to this laer aspect as socialisation
or cultural transfer (see §2.8.3). North’s view of culture is however closely connected to
that of sociological institutionalism, because he stresses its cognitive significance: “Cul-
ture provides a language-based conceptual framework for encoding and interpreting the
information that the senses are presenting to the brain” (North, 1990: 37).
The transmission of culture ensures that informal solutions for exchange problems
survive; social change may occur, but in the long run it will be accompanied by a persist-
ing undercurrent.
Informal institutions arise to coordinate repeated social interaction. Here again, North
distinguishes three variants:
1. First there may be exte nsions, elaborat ions and amendm ents of forma l r ules. These are the
‘unwrien laws’ or conventions that arise as a result of repeated interaction between
2. Then there are soc ially sanct ioned no rms of behaviour. North cites the example of a gentle-
man who is challenged to a duel to the death. The evening before he draws up a long
list of reasons for not taking part in the duel, with at the top the strong argument that
there is a fair risk that he will lose his life. Yet he still decides to participate in the duel:
a gentleman would suffer a serious loss of reputation if he avoids what in his circles is a
traditional manner of seling conflicts.
3. Finally, North points to the existence of internalised s tandards of conduct: ideas, ideolo-
gies and convictions which cannot always be reconciled with rational choice. Although
he does not mention him by name, Weber’s classical analysis in Die Protestantis che Ethik
und der Gei st des Kapitalismu s (1905) clearly resonates in North’s argument: “Effective tra-
ditions of hard work, honesty, and integrity simply lower the cost of transacting and
make possible complex, productive exchange” (North, 1990: 138).
In North’s view, in modern democratic societies informal institutions are important for
two reasons. In such a context the formal rules (political and judicial rules, economic
rules and contracts) usually are numerous and complex, but can never cover all possi-
      
ble circumstances. The rights established in the formal institutions therefore have to
be strengthened by informal constraints: conventions, social norms and internalised
standards of conduct that regulate the latitude of interpretation and action implied in
the formal rules.
In addition, the costs of expressing opinions in democracies are oen low, because
freedom of speech is a constitutional right which is ratified by the prevailing views of the
population, and because information is relatively freely available. These low costs mean
that the subjective preferences of individuals are more important determinants of behav-
iour than is the case under an authoritarian regime, where the preferences expressed
in public life have to toe the official line. The costs of expressing one’s opinion in such
a regime may be considerable: some statements are unlawful, those uering them are
regarded as social outcasts and moral reprobates, and censorship means a great deal of
effort is needed if individuals wish to make their views known to others.
2.2.3 Enforc ement
North (1990: 32-33) points to the importance of the ability to enforce desired behaviour for
both formal and informal institutions. This enforcement can be achieved through direct
reprisal by the injured party, through internalised codes of conduct, through sanctions
imposed by the community, or through the actions of a third party (oen the state). In
complex exchanges where no complete information is available, enforcement is needed
to achieve cooperative behaviour; without it the individual interests of the various actors
would predominate. The costs of enforcement form part of the transaction costs, and
reflect the uncertainties of the contract. They are a risk premium, whose amount depends
on the likelihood that the other party will fail to meet their obligations, and the costs that
this would entail for the first party. North comments that contracts in a profit-maximising
situation are self-enforcing’ when the rewards of complying with contracts are greater
than the costs. In his view this occurs in tribal societies and small communities, where
people have lots of information about each other and repeatedly engage in exchanges.
Here, the measurement costs of contracts are low, but the costs of cheating, avoiding
responsibility and opportunism are considerable. Formal contracts are not necessary;
norms of behaviour govern the exchange. Modern societies, however, are characterised by
large-scale impersonal exchanges. There are many valued aributes, the exchange oen
extends over a long period and is not repeated in the same form between the same actors.
In this situation the measurement costs are high, and without enforcement the advan-
tages of cheating would be much greater than the rewards of cooperative behaviour.
North clarifies this by calling on game theory. In the classic prisoner’s dilemma a
sub-optimum solution is oen chosen because the collective payoff is much lower if one
person cooperates (snitches his fellow prisoner) while the other does not. In a repeated
or iterative game, however, cooperation is a more evident choice, provided a number of
conditions are met: the information must be complete, and the exchange must continue
indefinitely (if people suspect that the game is finite, the anticipated final moment will
partly determine their choices). In realit y, such conditions are oen not met: the dura-
tion of the exchange is unknown, it involves several individuals who are not ver y well
acquainted, and the information possessed by the various parties differs widely. In this
situation enforcement and, in the case of non-compliance, punishment – by a third
part y is necessar y.
It is possible that voluntary organisations are the sole enforcement agent. How-
ever, North points out that in the context of impersonal exchanges in modern, mutually
dependent economies, the transaction costs for information acquisition and sanctioning
will quickly become too high. If the state acts as the enforcement agent, by contrast, it
can generate huge economies of scale. This then creates a new dilemma, however. On the
one hand a modern society cannot operate without formal ‘third-party enforcement’ by
the state; but on the other hand those who represent the state are also agents who seek
to maximise their utility: “Put simply, if the state has coercive force, then those who run
the state will use that force in their own interest at the expense of the rest of society”
(North, 1990: 59). This means that the state is not necessarily a neutral third party, which
can assess the value of aributes at lile or no expense, and which will automatically
ensure that those who break the rules pay so much compensation to injured parties that
defection is more costly than complying with agreements.
2.2. 4 Institut ional chang e and path dependence
North’s view of institutional change is an interesting mix of the rational choice, socio-
logical and historical institutionalism discussed earlier. He sees two driving forces: both
mutations in relative prices and altered actor’s preferences can lead to rule amendment.
He regards the former as the most important. This involves changes in the ratio of factor
costs (land/labour; labour/capital; land/capital), in the costs of information, and in civ il
and military technology. According to North, relative prices may change due to exogenous
shocks, such as the plague epidemic in Europe in the late Middle Ages, which radically
changed the price ratio between labour and land. However, he considers relative price
changes largely as a process driven by endogenous developments. The role fulfilled by
entrepreneurs is crucial in this respect. Over time they gain knowledge and experience,
and as a result their transaction costs fall, in line with the notion that a game proceeds
differently when it is played by professionals rather than amateurs. Therefore, if the pro-
fessionals gain the upper hand relative prices will change, and this will create pressure to
refine the rules of the game. Or to put it more precisely,
The process by whic h the entrepreneur acquires skills and k nowledge i s going to change rela-
tive prices by changing per ceived costs of me asurement and enforcement, and by altering per-
ceive d costs and benefits of ne w bargain s and contr acts (Nor th, 1990: 84).
If the relative prices change, this means it becomes aractive for some parties to recon-
sider existing contracts. Since these are embedded in the hierarchy of rules, this is oen
not possible without breaking higher rules or norms of behaviour. A disadvantaged party
may then decide to aempt to change the formal rules. This subsequently distorts the
equilibrium in the informal institutions, which are by nature embedded in the old formal
      
structure. A change in the relative prices thus may change the formal rules with the result
that behavioural norms, customs or traditions are gradually eroded and replaced by oth-
ers, which beer match the new formal institutions.
But North suggests that institutional change comes about not just because of changes in
relative prices. Changes in the preferences or ideas of actors are a second potential source
of change – in North’s view oen less important, but relatively autonomous nonetheless.
He illustrates this by pointing out that the abolition of slaver y in the United States cannot
be explained on the basis of changes in relative prices. At the time of the American Civil
War, slavery was still profitable in the Southern states, so there was no economic need for
abolition. North aributes the institutional change that took place aer the end of the
Civil War in 1865 among other things to the strong intellectual force of the anti-slavery
movement, combined with political democracy which enabled opponents of slavery to
express their ideas at lile personal cost. By contrast, the Southern slave masters had no
means at that time and in that structure to persuade the entire electorate to back their
point of view (North, 1990: 85). North argues that the causes of such changes in prefer-
ences, or ‘cultural evolution’, lie in random processes, learning mechanisms, and natural
selection. It is however also possible that preferences adapt in response to relative price
changes. To illustrate the laer North refers to the evolvement of family relationships
in response to changing relative prices of work and leisure time and the introduction of
North stresses that it is ultimately the actors who make and recreate the rules. The
existing institutions define an opportunit y structure for individuals, to which they will
gear their behaviour; and if their social success is too low according to the standards they
apply, they will try to change the formal or informal rules. North allocates a key role here
to political and economic entrepreneurs, an idea which corresponds with the view of
Eisenstadt cited earlier. To what extent they succeed in this depends on their cognitions
(formal knowledge, tacit knowledge and experience), their subjective norms and expec-
tations, their interests and negotiating power, and the degree to which they succeed in
competing with other interested parties in shaping the policy agenda and formal rules to
their own wishes. As their rationalit y is by definition bounded, this can lead to socially
sub-optimum solutions. Their bounded rationality also explains why entrepreneurs may
react differently to economic, social or technological changes.
As he also aaches importance to informal rules and to the perceptions and accumula-
tion of knowledge of entrepreneurs, for North institutional change theoretically is not
a linear process. Changes in preferences may lead to modification of informal institu-
tions, which in turn can prompt changes in the formal rules. There is also frequently
interference between the t wo: if formal institutions are changed radically, the informal
rules oen offer resistance to excessive reforms, because they still offer a solution to
the exchange problems of the actors. In these cases a new equilibrium may eventually
emerge, in which the formal institutions move in the direction of the ancien gime, and
the informal rules are modernised.
If the two driving forces – changes in prices and preferences are combined, this
offers a certain explanation for institutional variation in the short term, but still leaves
unclear why inefficient institutions are able to survive. Aer all, if certain countries or
organisations record lower achievements than others, the expectation would be that
they eventually succumb to the competition. They would necessarily have to opt for the
more efficient rules in the end, in order to prevent the best qualified part of the popula-
tion leaving the country or the business going bankrupt. The survival of sub-optimal
institutions can be explained by a third mechanism in North’s theory: the notion of path
dependence, which was discussed earlier under historical institutionalism. According to
North, path dependence arises as a result of the increasing added value of institutions –
as a result of high initial investments, learning effects, coordination effects and adaptive
expectations – in combination with imperfect markets. It is, however, not an automatic
or linear process (North, 1990: 98-99):
[There is no quest ion of ] a stor y of inev itabilit y in which the past neatly predict s the future.
[...] Path dependence is a w ay to nar row conceptually the choice set and link dec ision-making
through time. [...] Once a development path is set on a partic ular course, t he net work ex ter-
nalit ies, the learning process of the organizations, and the histor ically der ived subjective
modeling of the issues reinfor ce the cour se.
Path dependence is also one of the reasons that formal rules, when adopted by other
countries or organisations, do not function in the same way or may generate different
effects. German reunification provides an interesting example of this: the labour market
and social security institutions of the old Bundesländer were ‘exported’ to the new federal
states in the 1990s, but led to somewhat different results there (see Mars et al., 2002).
The foregoing makes clear that formal and informal institutions theoretically determine
the behaviour of actors, and thus influence the economic and social performance of
countries and organisations. For sociologists, the tenet ‘institutions maer’ is more self-
evident than for neo-classical economists, whose paradigm adherents of the economic
variant of rational choice institutionalism () oppose. The intellectual challenge is
rather to establish empirically the magnitude of the socio-economic impact of institutions
in diverging contexts. In this study this issue will be explored for specific types of rule
sets: those pertaining to systems of social securit y. Before tackling this empirical ques-
tion, however, it will be useful to elaborate the meaning of institutions more precisely,
and to specify what such a demarcation implies for social securit y. The first is covered
below, while the second is the subject of chapter 3.
2.3 The evolvem ent and impact of ins titutions: a f igural model
It is not simple to integrate the three variants of ‘new institutionalism’ discussed above,
partly because they have developed relatively autonomously. Here, building on the work
of North, an aempt is made to synthesise them into a general figural model. This is not
intended as a formal theoretical model, but rather as a heuristic aid, a flow chart of hypo-
      
thetical relationships that may be translated into empirical statements (cf. Lindenberg,
1983: 21-22). The model seeks to indicate what institutions are, how they come into being,
in what way they regulate interaction, and what consequences this has for both the actors
involved and societ y at large. Figure 2.1 illustrates the main outlines; the individual ele-
ments will be discussed in more detail in the following paragraphs.
The model distinguishes between factors at the collective and actor levels. Inst itutions are
the collectively defined rights and duties that aim to regulate behaviour, and their aached
conditions and possible sanctions. Formal institutions are rules promulgated or recognised
by the government; informal institutions are borne by groups or communities. In the hier-
archy of formal institutions, the meta-rules which establish the method of decision-making
are situated at the top. They give direction to the rules for gover nment product ion: goods and
services provided by the government (e.g. defence, education). They also influence the third
type of formal rules, namely enforcement by the government, as a part y above the parties,
of the rights, duties and mutual relations of private actors. This third-part y recognition forms
the framework for for mal contract s between private parties, the fourth type of formal insti-
tution. The contracts entered into by the government with private parties in its role as a
producer (tenders for public works, etc.) are a specific variant of this.
Among the informal institutions, values theoretically give direction to soc ial norms.
The former consist of general principles of action, the laer of specific behavioural rules
for actual situations. Unlike values, norms also contain the possibility of sanctions. Con-
ventions differ from social norms in that they do not have an explicit value component;
they specify correct behaviour in a neutral way. The content of such rules is arbitrary: the
main thing is that conventions coordinate the interaction within a group or community,
and confirm membership of it. Infor mal contrac ts are a fourth variant: contracts between
private parties that are not enforced by the government. Informal contracts are theoreti-
cally directed by social norms and conventions.
Formal and informal institutions are interrelated and therefore correspond to a cer-
tain extent. In principle, however, they do not determine each other in full: informal
rules do not automatically derive from formal rules, and the converse is equally untrue.
This also applies within the formal and informal rule hierarchies: the lower-level rules
bear a certain relationship to the higher-level rules, but do not automatically ensue from
them. Rather, the translation process is a maer of social consensus.
Institutions try to regulate action, and at this level two types of actors are distinguished.
In addition to individuals (natural persons) there are also corporate actors. These exist
where the rights and duties of such organisations (companies, associations, government
agencies) are separate from the individuals that form part of them; they are recognised as
independent legal body in their own right. Corporate actors are the result of the organ-
ising behaviour of other actors. They arise when individual or existing corporate actors
wish to achieve certain goals which they are unable to achieve by themselves. ‘Organis-
ing’ implies that existing actors accord a mandate and resources to a new or modified
corporate actor in order to realise such aims.