: The Generational Welfare Contract : Justice , Institutions and Outcomes

This groundbreaking book brings together perspectives from political philosophy and comparative social policy to discuss generational justice. Contributing new insights about the preconditions for ...

Simon Birnbaum, Tommy Ferrarini, Kenneth Nelson and Joakim Palme: The Generational Welfare Contract: Justice,

Institutions and Outcomes
Reviewed by Steffen Suur-Nuuja and Jörg Tremmel P olitics: Who Gets What, When, How is the famous title of a classic (1936) book by Harold D. Lasswell, and a cogent formulation of the key question of political science as an academic discipline.This question can be more specifically targeted and reformulated as: "Who, in the capacity of being a young, middle-aged or old person, gets which entitlements when from the welfare state?"This is the question, albeit framed a little differently (see below), at the heart of the book The Generational Welfare Contract: Justice, Institutions and Outcomes, written by Simon Birnbaum, Tommy Ferrarini, Kenneth Nelson and Joakim Palme.As one can see from the outset, the question is both normative (Who should get...?) and empirical (Who, de facto, gets...?).And the question is of a stubborn intractability as we all age and thus pass through different life spans.While Lasswell treated the "who" as one unified individual, the more refined question above treats it as the combination of intra-personal identities, that is: between our younger and older selves.Birnbaum et al. focus on three particular stages of the life course: childhood, working age and old age (9).For each of these three stages, the authors claim a specific "vulnerability" (2) that justifies the ascription of "social citizenship rights" (8-10) to individuals.The book adds to the growing part of the literature that is interested in intergenerational justice, but not in those abstract phenomena (such as the non-identity problem) that arise only if generations are treated as non-overlapping, non-coexisting entities.Birnbaum et al. understand "generations" as "age groups" and "cohorts".When the baby boomers born after the Second World War retire, they will not be replaced by cohorts of the same size.From this (uncontroversial) starting point, many other authors have questioned the long-term affordability of public programmes, such as health care and pensions, in their current extent.The proponents of an age crisis of the welfare state, namely its pension system, point to the massive impact (an "agequake", according to Wallace 2001) that the retirement of the baby boomers (if not postponed by a rising retirement age) will have for a welfare state's abilities to pay all kind of social expenditures (Preston 1984;Kotlikoff/ Burns 2012).In terms of benefiting from the welfare state, some scholars call the post-baby boomers a "disadvantaged" (Green 2017) or "precarious" (Bessant/Farthing/ Watts 2017) generation.The unwillingness of the "selfish" (Beckett 2010;Thomson 1991) baby-boomer generation (who have all the political clout at the ballot box) to give up their social entitlements (which are, according to this view, privileges rather than rights) and the dire prospects of the post-baby-boomer generation are two sides of the same coin.In a similar vein, but adding the assumption of an empowerment of the politically powerless post-baby boomers, the thesis of "generational storms" or "clashes" (Kotlikoff and Burns 2012) has been brought forward.Some authors have even hypothesised that the existing implicit generational contracts (such as pay-as-yougo pension systems) will be terminated soon by the youngsters if the ratio of public resources that go to the elderly relative to the amount of resources going to young people is constantly increasing.In contrast, Birnbaum, Ferrarini, Nelson and Palme proclaim the hypothesis that intergenerational welfare state contracts can lead to positive-sum solutions, and thus "it becomes an important task to identify and actively promote forms of intergenerational cooperation that enhance the welfare of all age groups" (3).This approach emphasises the potential to make all successive generations in a political community better off by mutually cooperating instead of relying solely on their own savings: "Allowing welfare states to redistribute resources between members of different generations as they pass through different age groups (or life stages) should be conceived as an arrangement for borrowing from our later selves in early stages of our selves, and to save for old age during the more economically active years in life in a way that effectively serves the long-term interests of all citizens" (20).This refers to Norman Daniels' prudential lifespan account as a concept for optimal allocation of resources during a lifetime.But is the step from intra-personal redistribution to intra-societal redistribution justified?The main counterargument is that different age groups can have different sizes -a problem obviously not affecting a single person.In line with the fathers of pay-as-you-go pension systems (such as Winfried Schreiber in Germany), the authors hold that one can reasonably expect the support of one's children if one has supported one's parents.This principle has also been dubbed indirect reciprocity in the literature (e.g.Tremmel 2009).The main idea is cogently summarised by a popular legend: " [A] boy in ancient times accompanies his father and his grandfather as they embark on a ritual journey intended to end with the grandfather's voluntary death, as he is no longer self-supportive.The boy takes pity on his grandfather and persuades his father to promise to support the old man until his natural death in exchange for a promise from the boy to do the same for his father when the time comes" (Lindh/Malmberg/Palme 2005: 470).But what if the (grand)children generation is not as numerous as their parents' generation?The authors concede that population ageing can lead to welfare state entrenchments.But they qualify: "The one-eyed focus on pension reform may severely distort conclusions about generational justice as it fails to recognize that increases in pension expenditures do not always come at the expense of younger generations and their access to adequate social protection" (33).Moreover, the authors contradict the "demography is destiny" thesis which postulates that ageing societies necessarily have to cut back their expenditure.Instead, they write: "It is an undeniable fact that social spending on old-age benefits has increased alongside population ageing" (32).The purpose of this book is to analyse how different welfare states respond to age-related social risks from a justice-based perspective, and if and under what conditions some countries perform better than others in promoting generational equity (2).To approach these matters, the authors want to bring together perspectives from two strands of academic research: political philosophy and comparative social policy.This is to respond to perceived shortcomings: "Despite the long tradition in normative political theory of debating principles of social justice and their practical implications, conceptual and theoretical discussions often remain at high levels of abstraction with limited reference to systematic empirical evidence" (4).Comparative welfare state research, for its part, has too often been reduced to crude analyses of social expenditures without having developed a coherent framework "that specifies central principles in welfare state program designs of particular relevance for analyses on generational justice" (4).This ambitious interdisciplinary approach is one of the strengths of the book.It is ambitious because it demands, to a certain degree, the authors to be familiar both with techniques of descriptive data analysis and regressions, and with the strands of the philosophical debate about generational justice.Such a combination of the normative and the empirical approaches is seldom found, but it has great potential for new insights.Normative statements should be "fleshed out" empirically whenever possible.For instance, the ongoing debate between protagonists of a "complete life course view" (Daniels 1988;Schokkaert/van Parijs 2003) and "relational equality" (McKerlie 2013; Bidadanure 2016) would strongly benefit from having their theories fleshed out with as much empirical evidence as possible.For readers who are not familiar with the debate: Daniels' assumption is that it is not prima facie problematic that at one given point in time different age groups receive an unequal treatment from the state.Let's assume that some countries spend much more on every elderly citizen than on every non-elderly one.Those who are old now were once young and those who are now young will some day be old.As long as the specific ratio value stays relatively stable over time, there is no inequality between people's complete lives, as everyone belongs in turn to each of the age groups.Such cases of "Spartan-childhoods for luxury-old-age" trade-offs (Vanhuysse/Tremmel 2018) are not per se signs of intergenerational injustice then.
Against this complete life course view, other authors have posited a "relational equality view", and argued that we should look at how people fare at each single stage of their lives independent of how they fare in terms of their lifetime as a whole.The fact that each generation X, while passing through their young age bracket, is dominated by the (then) old age group is not rendered fair by the fact that the same generation X will become the dominator (with regard to the succeeding generation Y) when they turn older and become members of the old age bracket themselves.Normative theories often abstain from real-life-contexts and are seldom operationalised.For instance, the relational equality theory implicitly suggests statements about the monetary level of an adequate minimum income, independent of age.Not making such statements explicitly is eschewing empirical tests of the hypothesised effects of such theories.Empirical research can and should inform normative debates.This is what the interdisciplinary work The Generational Welfare Contract tries to do.Both philosophers and social scientists who work on intergenerational justice should make the effort to read those parts of the book that are respectively less accessible to them.
Clearly structured, the introduction of the book is followed by three philosophical-theoretical chapters that discuss three perspectives on generational justice, thereby establishing the theoretical and normative framework of the empirical research that follows in the later chapters.In Chapter 3, and Chapter 4, the notion of a "balanced generational welfare contract" is defined, thereby at the same time conceptualising three kinds of welfare contracts that are unbalanced because they are "pro-child", "pro-work" or "pro-old".The authors here enter into a vivid debate (without citing details) about indicators for the alleged pro-elderly bias where ageing welfare states are more supportive of retirees than of other age-groups.One prominent indicator was developed by Vanhuysse (2013) with the elderly-bias indicator of social spending (EBiSS) (for previous such approaches, see e.g.Lynch 2006; Tepe/ Vanhuysse 2010).In an update, Vanhuysse/Tremmel (2018) state that within OECD countries currently Poland, Greece, Italy, Slovakia and the Czech Republic have the highest EBiSS levels: these states spend on average between 5.5 and 8.5 times as much on every elderly citizen as on every non-elderly one.With their own methodological approach, Birnbaum, Ferrarini, Nelson and Palme come to strikingly different results (52): Australia, Canada, Ireland, Italy, New Zealand, the UK and the US are the pro-old regimes of our times (whereas four countries display pro-work schemes and no countries pro-childhood schemes, the rest are "balanced").Unfortunately, the authors do not refer specifically to previous attempts to measure elderly-biased (or "pro-old") welfare schemes.Regarding their own methodology, Birnbaum et al. explain: "[W]e use income replacement in major age-related social insurance schemes to measure and analyse the generational structure of social citizenship.[...] For each age related social risk, entitlements are calculated net of taxes and expressed as percentages of an average production of worker's net wage.[...] For old-age risks we use the yearly pension benefit of two model families; a single retired person and a married retired couple.In both instances, the breadwinner is assumed to have a 40-year employment record.The non-working spouse only qualifies for a minimum pension, if applicable.Income replacement for old-age risks is an additive index of the net pension replacement rate of the two model families" (42-45).And regarding the dataset: "The database includes up to 47 countries and [...] we draw on data for 18 long-standing welfare democracies from 1960 to 2010 which is the most recent wave of data" (42).The distinction between "balanced" and "unbalanced" profiles is central for the rest of the book as "balanced profiles" is the independent variable in the regressions of Chapters 5-8 for welfare state outcomes, such as poverty or wellbeing.Therefore the methodological construction of "balanced profiles" (and thus "unbalanced profiles") deserves a closer look.Birnbaum et al. explain: "The cut-off used to determine whether social citizenship rights are balanced or not is of course to some extent arbitrary.We have for each country analysed differences in income replacement between the three age-related risk categories by calculating a straightforward statistical measure of dispersion.We decided that it is reasonable to categorize profiles of income replacement in age-related social insurance with a relative standard deviation below 20 percent as balanced" (50).If, on the other hand, one age group receives 20% more than the others, the scheme is defined as "unbalanced", as illustrated in Figure 1.
It now becomes clear what the authors mean when they speak about "positive-sum solutions": in countries where income replacement in social insurance is more evenly distributed across age-related social risks, the overall level of income replacement tends to be higher.A regression analysis ( 55) is conducted by the authors to lend evidence to the hypothesis that balanced schemes are causal for high overall levels of social expenditure, or, in the terminology of the authors, "high levels of social citizenship rights" (8).The next three chapters all deal with specific welfare state outcomes.As mentioned, balanced schemes (as constructed above) are used as the explanatory variable in a number of regressions that follow.In Chapter 5, the focus is on poverty, objectively measured.In Chapter 6, the focus shifts to wellbeing, subjectively assessed in the eye of the beholders.In Chapter 7, the analysis is complemented by a look on political and social trust; in Chapter 8, (un)employment comes to the fore.With regard to poverty, their regression analysis shows that countries with a balanced generational welfare contract have significant lower poverty risks over all three age groups (77).The intermediary variable is the high overall level of income replacement.Concerning subjective wellbeing (divided into life satisfaction and happiness), the analysis comes to the conclusion that balanced generational structures are related to high levels of subjective wellbeing (91).The intermediary variable, again, is the high overall level of income replacement.With regard to citizens' trust (in each other as well as in political institutions), there is a positive correlation between balanced schemes and trust, and the regression analysis corroborates the hypothesis that the former causes the latter.There are a few outliers, though.The Netherlands and Switzerland are two countries with the pro-work type scheme, but they nonetheless have levels of social and political trust that are on a par with countries in the balanced group (100).Being aware of conjectures in the literature that high levels of social expenditure might not foster high employment outcomes ("the dominating view in mainstream behavioral economics", 120), Birnbaum et al.'s study focuses on labour force participation as well.Their empirical results show that "(un)employment appears to be largely unrelated to the ways in which countries have organized their generational welfare contracts" (120).In the concluding Chapter 10, the Swedish/Finnish author team reiterates the main findings and summarise: "The story that we are telling in this book thus clearly diverges from the narrative of an unavoidable generational war in social policymaking" (141).
The book has already received a lot of praise.Gøsta Esping-Andersen (Universitat Pompeu Fabra) has called it "arguably the single most important welfare state study in our times" (front cover).The Generational Welfare Contract is indeed an innovative, sensible and topical work.There is no room here to discuss the proclaimed causalities between "balanced generational contracts" and a number of welfare state outcomes in detail.The most important contribution of this study, in our view, is one aspect that went largely unnoticed or is at least not particularly emphasised by the Swedish/Finnish author team: they have laid out a framework to further empirically test one vividly debated theory, the "selfish baby boomers" theory.This needs to be explained: as mentioned, two different concepts of "generation" are relevant with regard to public spending (for an illuminating visualisation, see Vanhuysse/ Tremmel 2018).First, when we want to evaluate intergenerational justice over complete lives, we need the concept of (birth) cohorts.These are groups of people who were born in the same year or narrow range of years.Cohort members, by virtue of ageing together, experience distinct public policies but also external events (such Figure 1: The construction of the "balanced" scheme and an "unbalanced" scheme (here: pro-old) in comparison as deep recessions for the worse or technological progress for the better).Second, when we want to make a snapshot analysis of intergenerational justice at a given moment in time, we need to look at age groups which are people of the same (narrow) age bracket at a particular moment.Age group members find themselves in the same stage in the lifecycle, which is politically relevant because public policies tend to institutionalise the life course, proscribing and inhibiting certain behaviours.As mentioned, many scholars (with Daniels) hold that it is not prima facie problematic that at one given point in time different age groups receive an unequal treatment from the state.But if such inequalities are perpetuated across different birth cohorts over the entire life cycle, then we do end up with intergenerational inequities.While fairness between age groups can involve unequal benefit treatment in different life stages, fairness between birth cohorts implies enjoying approximate equality in benefit ratios.To make such statements, empirical data must add up to a longitudinal series of the "snapshots", mentioned above, for as many decades as possible.Using the EBiSS as an indicator, Vanhuysse/Tremmel 2018 derive from these considerations the conceptual statement that if those countries that have for instance a high EBiSS in 2010 had a low one 40 years ago, age-group inequality would turn into cohort injustice, for this meant nothing else than people that were in their twenties and thirties in 1970 (that is: forty years ago) profited a lot from state benefits in all stages of their lives.One of the few empirical studies on how different generations have fared under the social welfare policies of governments since the 1930s is David Thomson's (1991), who argues that in New Zealand "the big winners […] have been […] those born between about 1920 and 1945.Throughout their lives they will make contributions which cover only a fraction of the benefits" (Thomson 1991: 3).Recent studies (Chauvel/Schröder 2014; see also Chauvel 2010) have shown some empirical evidence that in countries in Southern Europe such as Spain, Italy, and France, the baby-boomer generation born after the Second World War has been significantly better off in terms of post-tax-and-transfer disposable income than cohorts born both beforehand and afterwards.In Birnbaum et al.'s methodology, the childhood bar would need to be highest in 1960, the working-age bar highest in 1980 and the old-age bar highest in 2010 to justify the "selfish baby boomers" hypothesis.
Birnbaum, Ferrarini, Nelson and Palme's empirical analyses show diverging results for the 18 countries they have decided to look at (46-50).Some countries, such as Canada or Ireland, have an exchange between the pro-work bar and the pro-old bar between 1980 and 2010 in favour of the ageing baby boomers, but a clear trend is not visible (and was not the focus of the Swedish/Finnish author team).A consolidated account of the 18 OECD countries depicts that in 1960, childhood entitlements were quite low (around 20%) while working-age and old-age entitlements were about the same and about 40%.In 1980, the relative height of all bars is still the same but they all are higher.In 2010, all bars have approximately the same height, therefore forming what the authors call a "balanced" generational welfare contract (see Figure 3).This might not be enough to repudiate the "selfish baby boomers" hypothesis (which in some of its variations is not focused on the past, but on the years when the baby boomers retire, that is from 2025 on), but it lays out the path for further research.All else being equal, interdisciplinary approaches are more ambitious but also often more promising than mono-disciplinary ones.And in the case of The Generational Welfare Contract, the authors have lived up to the promise.3: Income replacement in social insurance for three age-related social risks (average for 18 OECD countries) (Birnbaum et al. 2017: 47)

Figure 2 :
Figure 2: Conceptualisation of the "selfish baby boomers" hypothesis, applying Birnbaum et al.'s methodological framework

Figure
Figure3: Income replacement in social insurance for three age-related social risks (average for 18 OECD countries)(Birnbaum et al. 2017: 47)