Objective or Subjective Well-Being?

Abstract

This chapter focuses on the objective-subjective debate in defining and measuring child well-being. It starts by outlining the underlying philosophical perspectives and how this is reflected practically in conceptual and policy frameworks for child well-being. Particular attention is paid to the concepts of outcomes and risk and protective factors. This is followed by an analysis of approaches to measuring objective and subjective well-being. There is a focus on the use of standardized measures, illustrated by a brief case study of the application of such an approach. Next, the chapter considers some of the connections and contradictions between objective and subjective accounts of children’s lives. It then considers the functions of objective and subjective well-being data in policy and practice and how these can be enhanced. The final section includes observations about where the debate on objective vs. subjective well-being is heading and what this means for research, policy, and practice.

Parts of this chapter have previously been published as Axford (2008, 2012), Axford and Hobbs (2010), Axford et al. (2012a), and Hobbs et al. (2010).

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

In recent years, the view that rising material prosperity and market individualism leads to greater well-being, including for children, has come under fire (Jordan 2006). Evidence emerged that aspects of child mental health in some developed countries had declined over a prolonged period of economic growth (Collishaw et al. 2004) and that child well-being in some rich countries was often no better, and even worse, than in less well-off countries (Bradshaw et al. 2007; Bradshaw and Richardson 2009). This fostered a robust debate about the society in which children are growing up and what a “good life” for children looks like (Rose and Rowlands 2010). The distinction between objective and subjective well-being has been a major fault line in this debate.

Traditionally, child well-being indicators have been drawn from existing administrative data sources and have been more objective and determined by adults’ views of children. This is reflected in major international and national monitoring efforts, such as KIDSCOUNT in the USA and the UN State of the Child reports. Latterly, there have been more efforts to ask children about their subjective well-being. Some approaches ask how they feel about the discrete aspects of their lives, others try to estimate overall, or “global,” subjective well-being, while others ask children about what is important to them and use the responses to inform measures (Fattore et al. 2009). This growing interest in subjective well-being (Moore and Keyes 2003) stems, in part, from the apparent discord between measures of subjective and objective well-being (Layard 2010; Oswald and Wu 2010) and the case for measuring both in order to gain a deeper and more holistic understanding of children’s well-being (Cummins 1997; Diener and Suh 1997). It has been galvanized by the UN Convention on the Rights of the Child, which emphasizes the importance of adults listening to the opinions of children and taking these opinions into account. In addition, it is increasingly recognized by policy makers that children with poor subjective well-being are especially vulnerable and may be a risk to others (Rees et al. 2012). So what is meant by “objective” and “subjective” well-being?

2 Defining Objective and Subjective Well-Being

This chapter is concerned mainly with quantitative data on the state of children’s objective and subjective well-being. Objective well-being covers different dimensions of children’s health and development – behavior, emotions, attainment, and so on – and, to some extent, the factors that impinge on these, such as housing, parenting, environment, and socioeconomic situation. Subjective well-being concerns how children and young people assess their lives, in particular how satisfied or happy they are with aspects of their lives or their lives overall.

It is helpful to conceptualize children’s objective well-being in terms of outcomes and risk and protective factors. Outcomes refer to aspects of children’s health and development that services may seek to improve, which can be subdivided into dimensions relating to, for example, behavior, emotional well-being, relationships, physical health, and educational skills and attainment. In developed countries like the UK, there has been far more attention in children’s services to outputs (i.e., administrative or service indicators) than to outcomes (i.e., intrinsic aspects of children’s health and development). This has started to shift in recent years, however, as indicated by the emergence of national “outcome frameworks,” indices, and sets of indicators (Jordan 2006; Bradshaw 2011).

Risk and protective factors refer to potential influences on children’s health and development, such as parenting behavior, maltreatment, the home environment, bullying, norms within the school or community, and economic influences (Little et al. 2004; Catalano and Hawkins 1996). Risk factors refer to characteristics, experiences, or contexts that increase the likelihood of specified negative outcomes. Harsh or ineffective parenting skills, for example, contribute to child behavior problems. Protective factors are characteristics, experiences of contexts that buffer children from negative outcomes in the context of specified risks (Little et al. 2004). A supportive adult, for instance, protects children who have been maltreated from various maladaptive outcomes. Risk and protective factors help us understand how problems develop in children’s lives and how to prevent or address them and may also be used to ensure that interventions are effectively targeted to local patterns of risk and protection (Hawkins et al. 2008).

There is a vast literature on subjective well-being and related concepts, including happiness, life satisfaction, self-esteem, and positive affect (Bradshaw and Keung 2011a). Essentially, subjective well-being refers to how individuals appraise their own satisfaction with life (Cummins 1997; Diener 2009; Rees et al. 2010). Put another way, it is a term used to describe people’s assessments of, or happiness with, their lives as a whole (Rees 2012). It is built up from understanding happiness and satisfaction in discrete life domains, such as home life, friendships, school, and material wealth. The concept is distinct from more objective measures of children’s health and development, such as behavior or physical or mental health symptomatology (Lucas et al. 1996). Instead, it concerns the subjective interpretation and value that children place on discrete aspects of their own life and their life as a whole, irrespective of their objective circumstances (Diener 2009).

2.1 The Good Life

The definition and operationalization of objective and subjective well-being are underpinned by responses to the question “What is the good life?” These may be categorized in terms of three broad perspectives, the first of which focuses on utility – maximizing happiness and satisfying people’s preferences (Cohen 1993). In this formulation, well-being is a person’s state of mind, which in turn depends on how far their desires are fulfilled and whether or not they are able to enjoy chosen activities and relationships (Dasgupta 1993). It should therefore be evaluated subjectively and is essentially “bottom-up.” This can be subdivided into hedonistic accounts of subjective well-being and desire theories: the former focus on maximizing pleasure and feelings of happiness while also minimizing pain (Eid and Larsen 2008), while the latter are more concerned with achieving desire preference and fulfillment on the assumption that this “always or nearly always brings satisfaction” (Gasper 2004, p. 7).

There are various criticisms of the utilitarian position. One is that it legitimizes the pleasure a person derives from hurting others, counting it as equal to other satisfactions in the calculus of what is good (Rawls 1971). It also risks overlooking actual deficits if the victim has a positive disposition. Thus, someone whose affliction is countered by a phlegmatic outlook or limited aspirations would not be found wanting by the utilitarian (Cohen 1993). Utility is hard to quantify, partly because it can be distorted by the experience of and adaptation to oppression or deprivation (Sen 1993).

The second approach to defining “the good life” therefore involves identifying the major ingredients of utility – the substantive goods that always make life better (Scanlon 1993). This perspective is arguably more objective; indeed, efforts to capture these ingredients are referred to as “objective list theories” because they “list items constituting well-being that consist neither merely in pleasurable experience nor in desire satisfaction” (Crisp 2008). Rawls (1971, pp. 60–65), for example, argued that well-being should be defined in terms of an individual’s bundle of primary goods. These are the things that every rational person is presumed to want, including liberties, opportunities, income, and other factors that contribute to self-respect. Similarly, the basic needs approach to eliminating destitution defines well-being in terms of commodities like food, shelter, clothing, and healthcare (Sen 1993). This approach stresses the human capacity for deliberation and doing and can be judged externally or objectively (Dasgupta 1993). It is top-down in that it involves identifying at a level of philosophical generality the rules that generate the good life.

Critiques of this perspective target what Sen (1982, p. 366) calls “commodity fetishism” – interpreting goods as the embodiment of advantage. Rigid lists of items overlook the fact that different people derive different benefit from the same good. Food intake and nutritional achievement, for example, vary with culture, climate, and a person’s metabolism and body size (Drèze and Sen 1989). People also have different tastes; what one person enjoys another person dislikes. The focus should therefore be less on ownership of primary goods than on what these goods do to people. As Sen (1985, p. 28) put it, “a person’s well-being is not really a matter of how rich he or she is . . . commodity command is a means to the end of well-being, but can scarcely be called an end in itself.”

In the third interpretation of “the good life,” Sen (1982, p. 30) christens the missing link between goods and utility “capabilities.” He conceives life as a combination of beings and doings (“functionings”), ranging from the elementary – being well-nourished, sheltered, clothed, and disease-free – to the more complex, such as having self-respect and participating in community life. A person’s capability refers to the various alternative combinations of functionings from which he or she can choose (Nussbaum and Sen 1993, p. 3). The stress is on what people are free to do and be – the ability to lead a life of one’s choosing (Sen 1999). Utility is evidence of that capability, while goods are the source of the capability.

Some commentators have attempted to put meat on these bones. Nussbaum (1995), for instance, identifies a number of basic capabilities, such as avoiding unnecessary and non-beneficial pain, being able to laugh and play, and being able to have attachments to things and persons. The theory of human need elaborated by Doyal and Gough (1991) also fits here. It identifies physical health and autonomy as the two basic universal needs and then lists 11 intermediate needs that contribute always and everywhere to meeting basic needs, including adequate nutritional food and water, a nonhazardous physical environment, appropriate health care, security in childhood, significant primary relationships, and basic education. For items to be universal, they must contribute to well-being, irrespective of a person’s environment, constitution, moral beliefs, and preferences (Scanlon 1993).

Although difficult to operationalize, Sen’s contribution is important because it demonstrates that neither the utility nor the substantive good approaches in isolation capture well-being: “a comparison of – say – malnutrition of different people is neither a comparison of foods they had consumed (they could consume the same amount of food and still have different levels of malnourishment), nor a comparison of utility (they could be equally malnourished but their levels of satisfaction or desire fulfillment could still be quite different)” (Sen 1982, p. 30). Rogerson (1995) describes the relationship in terms of an internal psychological-physiological mechanism that produces a sense of happiness and satisfaction, and those external conditions, for example, health or the environment, which trigger this mechanism. In the same vein, Lane (1996) uses the formula QoL = f (QC, QP), where QoL is quality of life, QC is quality of objective conditions, and QP is a sense of subjective well-being combined with personal development, learning, and growth.

2.2 Objective and Subjective Well-Being

Well-being is a holistic concept in which people are seen as biopsychosocial beings whose well-being hinges on efficiency of body, mind, and spirit (Bowling 1997). The World Health Organization QoL Group (1993, p. 30) considered the closely related concept of quality of life to be “a broad ranging concept affected in a complex way by the person’s physical health, psychological state, level of independence, social relationships, and their relationships to salient features in their environment.” Accordingly, most definitions of well-being break life into several domains and incorporate similar components. Approaches range from the simple to the elaborate, for example, Rowan (1980) distinguished between environmental and personal factors, while Flanagan (1978) grouped 15 categories under five headings: physical and material well-being; relations with other people; social, community, and civic activities; personal development and fulfillment; and recreation. Others are more philosophical, for instance, Renwick and Brown (1996) identify three areas of life in which all people strive for quality: being, the basic aspects of individuals; belonging, the fit between individuals and their environment; and becoming, purposeful activities to realize one’s goals.

Typical factors can be grouped into six dimensions (Axford 2008):

• 1. Physical well-being, including functioning, self-care, and the absence of disease

• 2. Material welfare, notably decent living standards

• 3. Social relations, in particular support from family and friends, involvement in cultural and leisure activities, and the development of reciprocal relationships

• 4. Environment, which encompasses everything from attractive architecture and green spaces to global security and a sense of neighborliness, together with the notion of “fit” – the idea that well-being is a function of the discrepancy between an individual’s objective circumstances and their aspirations

• 5. Psychological health, for example, happiness and life satisfaction

• 6. Spirituality, by which is meant personal development and fulfillment, growth, and mastery

These dimensions may broadly be applied to children and encompass both objective (1–4) and subjective (5–6) perspectives. The importance of the subjective perspective is given added weight by an understanding of the importance of the concept of “self-determination.”

Self-determination refers to a person’s capacity to exercise choice in forming and pursuing a coherent life plan. Conventionally, it has been regarded as independent of – and even antagonistic toward – well-being: for example, a patient who refuses the prescribed treatment is arguably acting against her best interests. An alternative perspective is that physical well-being and self-determination can contradict each other, but that they cannot be in opposition to good well-being overall (Brock 1993). This is because the freedom to choose makes a significant contribution to people’s happiness and desire satisfaction. The implication is that well-being requires the formulation of desires, hopes, and plans for the future; it is life lived from the inside. As such, the subjective perspective is clearly very important.

Another way of putting this is to say that since well-being comprises inputs and outputs, judgments about it are necessarily composed of observable facts – for example, regarding a person’s functional capabilities and environment – which can be apprehended by an external social group and the person’s valuation of those facts. The two aspects are captured in Renwick and Brown’s (1996, p. 80) definition of quality of life as “the degree to which a person enjoys the important possibilities in his or her life.” It is how objective circumstances are perceived relative to one’s outlook and time of life that determines how much pleasure one enjoys.

Combining objective circumstances and subjective appraisal in this way is important because, as will be illustrated later in this chapter, they can be at odds. For example, in the USA, the steady increase in material wealth since 1957 has been accompanied by declining levels of happiness (Lane 2000), and there seems to be a point at which the correlation between GDP per capita and well-being breaks down (Jordan 2008; Searle 2008). Similar disjunctions are evident in relation to physical health. Patients detect significant changes in their health that clinical measures miss (Fitzpatrick 1996), with the result that, for instance, a clinical measure of airways obstruction, such as peak expiratory flow, may not reveal a problem, despite the patient’s experience of breathlessness when walking up stairs (Hopkins 1992). Conversely, an important study of seriously debilitated people found that fewer than half were dissatisfied with their health: a handful (6 %) were “completely satisfied” (Campbell et al. 1976). Other medical research reveals similar patterns. Indeed, research reveals perceived positive consequences of disability and chronic disease, including improved interpersonal relationships and the discovery of new interests (Sodergren and Hyland 2000).

There are several reasons for the objective-subjective disjunction. First, objective measures may focus on factors that are important to the researcher or professional but of little consequence to the subject (Raphael 1996). Second, vulnerable populations, especially the frail elderly and disabled, can exhibit a “false awareness” borne of a lack of hope about attaining goals and a consequent lowering of expectations (Raphael 1996). Thus, many report good subjective well-being despite their objectively inferior conditions. Third, prior experience mediates the relationship between objective circumstances and subjective appraisal. Offer (1996) reports that the labor camp inmate of Soviet dissident Solzhenitsyn would consider it a good day in which nothing worse than the usual privations occurred. Fourth, people adapt or become accustomed to their new situation. When the onset of physical impairment makes activities previously associated with optimal experience unavailable, the individual concerned will often find alternative sources of enjoyment and engagement – a process called “transformation of optimal experience” (Delle Fave and Massimini 2005). Equally, the perceived benefits of comforts and wealth – or other good objective circumstances – can fade. For example, people may find that there are problems associated with better material circumstances – such as a greater need to protect their wealth (Searle 2008). Lastly, personal resources and traits bear on how an individual experiences their objective circumstances. Motivation, cognitive ability, patience, a positive temperament or affective style, sociability, a meaningful view of the world, and a sense of control are all relevant here (Day and Jankey 1996; Huppert 2005). They can enable people to overcome disadvantage and maintain a cheerful disposition; for example, some patients with crippling injuries adjust their aspirations and find happiness despite their newly limited functional capacities (Fitzpatrick 1996). Other individuals, perhaps lacking in some of the resources and traits identified, struggle to exploit favorable circumstances and remain miserable. As Lane (1996, p. 273) put it, “Opportunities in the environment contribute to quality of life only if there are matching receptive properties in the person involved.” The ability to enjoy fulfilling work, for instance, requires some degree of cognitive understanding and motivation, while the benefits of friendship come only to those with at least a modicum of self-esteem and ease of interpersonal relations.

Taken together, these points suggest that neither the objective nor the subjective approach to capturing well-being is adequate in isolation. Some commentators have therefore attempted to bring the objective and subjective together by focusing on the perceived gap between the “actual self,” what an individual can do, and the “ideal self,” what that person would like to be able to do (Calman 1984). Taking this a step further, Michalos’ (1986) “multiple discrepancies theory” holds that well-being is the product of comparing one’s current life to various standards – other people, the past, expectations, an ideal, and so forth. This approach is potentially helpful given the tendency for research often to investigate elements rather than the totality of well-being (Day and Jankey 1996, p. 50), although some research casts doubt on the ability of multiple discrepancies theory to explain differences in subjective well-being (Blore et al. 2011).

3 Measuring Objective and Subjective Well-Being

3.1 Social and Health Indicators

The measurement of well-being owes much to the social indicators movement and research into health-related quality of life. A plethora of social indicators for measuring well-being emerged in the 1960s in response to critiques of economic measures, notably GDP (Searle 2008). Most of them deal with objective external conditions – health, housing, education, transport, recreation facilities, civil liberties, and so forth – and have been mapped at different geographic levels, especially in the USA and western Europe. Many have been collated in statistical series produced by national statistics agencies, such as Social Trends in the UK, or international organizations, such as the UNDP Human Development Reports.

In relation to children, data have been compiled under the auspices of bodies such as UNICEF, whose annual State of the World’s Children reports review basic indicators on children’s development (e.g., infant mortality, school enrolment, percentage immunizations). In a similar vein, Micklewright and Stewart (1999) compiled Eurobarometer data for 15–19-year-olds in EU countries against indicators such as health, education, material well-being, social inclusion, and life satisfaction. These and similar developments were significant given that children were traditionally invisible in statistical returns (Saporiti 1999; Roche 2001). More recent initiatives include national frameworks for conceptualizing and measuring child well-being (e.g., Brooks et al. 2010) or cross-national comparisons sponsored by international bodies (e.g., UNICEF Innocenti Research Centre 2010).

The measurement of child well-being has developed in several ways in recent years (Ben-Arieh 2010). For example, there is a greater focus on subjective as well as objective well-being. This includes paying greater attention to the child’s perspective rather than relying on existing indicators or the views of parents or teachers and involves incorporating new domains of well-being, such as children’s civic involvement (Ben-Arieh 2010). This is reflected, for example, by the inclusion of questions about subjective well-being in the UK national household surveys (Tinkler and Hicks 2011). There has also been a move away from measures of survival and deficiency toward positive indicators such as satisfaction, neighborhood facilities, and participation in activities (Ben-Arieh 1997, 2010; Ben-Arieh et al. 2001). This is important, because the most widely used indicators – calories per head, literacy, access to sanitation facilities, and so forth – are better at capturing levels of basic need satisfaction rather than well-being (Gordon and Spicker 1999). There has also been more emphasis on positive well-being in recognition that “the absence of problems or failures does not necessarily indicate proper growth or success” (Ben-Arieh 2010, p. 133).

Meanwhile, research on well-being in medicine expanded rapidly from the 1970s onward, particularly in the fields of nursing, gerontology, health promotion, and rehabilitation (Renwick et al. 1996). It is framed in terms of health-related quality of life and has included the development of a number of generic and disease-specific well-being measures in pediatrics (Eiser et al. 2000a; Skevington and Gillison 2006). Most of this research have been concerned with evaluating posttreatment outcomes (Brown et al. 1996). Just as GDP had been deemed inadequate for measuring social progress, so standard health indicators such as increased longevity had become unsuitable for assessing modern medical interventions dedicated to relieving symptoms, improving mental health, and restoring functioning (Najman and Levine 1981). Standard health status indices include survival, morbidity, or service use but generate little information about the 80–90 % of the population who are healthy by such definitions (Bowling 1997). In addition, the capacity afforded by technological advances to manage life-threatening conditions needed to be weighed against often painful and distressing treatments, not to mention the diminished quality of years of life gained. For example, 80 % of children with acute lymphoblastic leukemia – the most common form of childhood cancer – can expect 5 years of disease-free functioning following treatment, but many suffer fatigue, cardiac damage, and social and psychological maladjustment (Vance et al. 2001).

Most health-related quality of life research uses psychometric instruments and questionnaires, of which there are many. For example, Bowling (1997) describes 58 scales for capturing functional ability, broader health status, psychological well-being, social relations and support, and life satisfaction and morale. Health-related instruments tend to cover one or more of three dimensions (Fitzpatrick 1996). First, they address functions of a biological, mental, physical, and social nature, all of which are necessary for the pursuit of relatively full and complete life plans. These include rest, eating, work, recreation, and home management. Second, they cover self-determination, as defined earlier in this chapter. And third, they assess patients’ conscious experience, in other words subjective reflections ranging from abject misery to happiness. The mix of objective and subjective approaches is again evident.

Instruments generally list items against life domains and elicit responses to each on set scales. Some are self-administered, while others are completed by a third person such as a carer or professional; for the under 8s, the vast majority of measures are completed by carers or doctors on children’s behalf (Eiser et al. 2000a). Results are calculated in different ways. In the case of an index, item scores are summed – often following weighting – to form a composite score, otherwise the results form a profile. The robustness of a measure is assessed in terms of validity, whether it captures the underlying attribute, and reliability, whether it consistently produces the same results if applied to the same subjects at different times when no change is evident.

These developments in well-being measurement were not isolated to health. As will be seen later in this chapter, numerous measures of multiple aspects of objective child well-being have been developed. In addition, as interest in children’s subjective well-being has grown, so too has the technology and number of instruments available to assess it (Diener et al. 1999; Moore and Keyes 2003). Subjective well-being is typically considered in terms of overall satisfaction with or quality of life, measured by tools such as Cantril’s ladder (This approach is used in the Health Behaviour of School-aged Children survey in the UK, as follows: “Here is a picture of a ladder. The top of the ladder, 10, is the best possible life for you and the bottom, 0, is worst possible life for you. In general, where on the ladder do you feel you stand at the moment?” Scores of 6 or more are considered to represent positive life satisfaction.) or Heubner’s life satisfaction scales (e.g., Huebner et al. 1998). In turn, overall subjective well-being is often disaggregated by exploring a number of sub-domains. These typically include constructs such as emotional well-being, physical health, family relationships, peer relations, school life, material or economic wealth, appearance, and autonomy (Bradshaw et al. 2010; Cummins 1997; Diener et al. 1998; Ravens-Sieberer et al. 2001; Rees et al. 2010; Ryff and Keyes 1995). Good examples of multidimensional measures of children’s subjective well-being include the Good Childhood Index (Rees et al. 2010), the Kidscreen instruments (Ravens-Sieberer et al. 2001; The KIDSCREEN Group Europe 2006), and the Child Well-being Index for School Children (Cummins and Lau 2004).

Instruments to measure child well-being are typically subjected to at least four general criticisms. The first concerns the choice of indicators, which can appear to be arbitrary and, beyond their immediate context, occasionally bizarre. For example, Najman and Levine (1981) identified studies that used height, weight, or desire to work, none of which are essential attributes or determinants of well-being. Second, instruments are criticized for being too expansive or too reductive. Some measures emphasize particular aspects of well-being rather than take a holistic perspective. Many health-related quality of life measures are disease specific rather than generic – for example, the Pediatric Cancer Quality of Life-32 (Vance et al. 2001). Other measures attempt to squeeze human reality into a single dimension such as economic status or life satisfaction (Rogerson 1995; ICPQOL 1996). Third, there are doubts about the validity of some measures. For example, self-reported physical functioning would be expected to concur with objective mobility tests (criterion validity), and scores for psychological well-being should arguably correlate with self-report data from recognized depression tests (convergent validity). These indicators should not correlate with factors unrelated to general health, such as height (discriminant validity). Many indices fall short of such standards (Muldoon et al. 1998). Fourth, instruments often struggle to capture an individual’s subjective experience. Sometimes only objective indicators are included. In other cases the content and weighting of measures will suit some people better than others (Bowling 1997) – for example, people with long-standing illnesses place particularly high value on mobility, communication, and self-care, but other people might take these for granted. A related argument is that children’s well-being may best be ascertained by eliciting from them the indicators as well as the responses (MacGillivray and Zadek 1995, p. 18). Finally, many instruments use third parties as their source, an issue elaborated in the following section.

3.2 Subject or Third-Party Response

Subject or third-party response has been debated at length in relation to health-related quality of life. Some commentators argue that it is vital that subjects express their views when measuring their well-being due to discrepancies with proxy respondents (Muldoon et al. 1998). For example, patient and doctor ratings of posttreatment well-being are frequently at odds (e.g., Slevin et al. 1988), and adults often struggle to reflect children’s views accurately; for instance there are significant differences between parent and child accounts using the Childhood Asthma Questionnaire (Eiser et al. 2000b). Various factors account for such divergence. One is knowledge: chronically ill children, for example, tend to evaluate their well-being more positively than their parents do, perhaps because the latter are more aware of or concerned about likely future effects such as missed schooling (Vance et al. 2001). Underlying values, preferences and development may also play a role, with children known to have different thought patterns and priorities than adults – a product of their earlier stage of cognitive development and generally more limited life experience (Hill and Tisdall 1997). Different individuals also have different baselines. In particular, third parties base their judgments on intuition concerning the transition into ill-health, forgetting that to some extent patients get accustomed to being ill and adjust their expectations accordingly (Fitzpatrick 1996). Lastly, extraneous factors impinge on third-party valuations, for example, parents’ stress levels have some bearing on how far their views tally with their child’s (Vance et al. 2001).

Lane (1996) advances a counterview, namely, that an assessment of an individual’s well-being need not be made by that individual. There are at least three reasons for this. First, it may not be appropriate to seek the subject’s views, perhaps because neurological or psychological dysfunction limits their ability to report accurately (Muldoon et al. 1998). Second, self-report may be misleading. Some subjects give socially desirable responses – to please their caregiver, for example – or hesitate from expressing dissatisfaction for fear of losing a service (Raphael 1996). Different people mean different things by “very satisfied,” or the same individual responds differently depending on mood or time of day (Day and Jankey 1996). Third, it may prove difficult methodologically to elicit the subject’s view. Administering self-report well-being instruments to younger children, for instance, may encounter challenges such as short attention span, less-developed language and reading skills, and limited ability to distinguish between points on a response scale (e.g., Spieth and Harris 1996).

Nevertheless, there has also been a strong move in research and policy toward treating children as active participants in society rather than passive subjects waiting to become adults (James and Prout 1997). This has been reinforced by research showing that parents do not always know what their children do with their time (Funk et al. 1999) or how they feel (Gottlieb and Bronstein 1996; Shek 1998; Sweeting 2001), especially in adolescence. Moreover, children often have a distinctive standpoint on issues, particularly when it comes to their own lives (Mayall 2002). This shift has been reflected in the child indicators movement, with a growing acknowledgement of the importance of obtaining data directly from children (Ben-Arieh 2010). These developments are reinforced by evidence that even young children can process fairly complex tasks competently and compare their own experience with that of others (Eiser et al. 2000a) – particularly where concrete examples, pictures, and other props are used.

3.3 Toward Greater Precision

As demonstrated, there are a plethora of indicators of child well-being. However, there remains a danger that what exists is often little more than collection of readily available figures; in other words, they are not informed by theory. In order to address this issue, recent research commissioned by the Annie E. Casey Foundation, as part of the Evidence2Success initiative, sought to develop a suite of conceptually grounded instruments to inform the delivery of evidence-based prevention and early intervention (Hobbs and Elliott 2013). Within this new framework, “key developmental outcomes” are defined as those outcomes that are most critical for children’s subsequent positive development across five broad areas: behavior, emotions, relationships, physical health, and educational skills and attainment. If children are not achieving these outcomes at particular stages of development, there is a strong likelihood that their future health and development will suffer. For example, if children cannot read at the expected level when aged 8–9 years, this is likely to have a negative effect on their subsequent educational skills and attainment (National Research Council 1998; Annie E. Casey Foundation 2010), just as antisocial behavior in early childhood is likely to persist into adolescence and early adulthood (Moffitt 1993). If children do not meet key developmental outcomes in one area, such as reading, this is also likely to be detrimental for outcomes in other areas, such as behavior (Miles and Stipek 2006).

Within this framework, key developmental outcomes, alongside a range of associated risk and protective factors, have been determined based on consistent and compelling evidence that they are predictive of children’s subsequent health and development (Hobbs and Elliott 2013). At least two multivariate longitudinal studies consistently demonstrating a strong prediction of impairment to subsequent health and development are required to demonstrate such an association. There is strong evidence, for example, for the link between children’s behavior at early childhood and their subsequent engagement in antisocial or delinquent behavior, substance misuse, educational skills and attainment, peer relationships, emotional well-being, and, to a lesser extent, physical health (Moffitt 1993; Fergusson et al. 2005; Colman et al. 2009; Kosterman et al. 2009; Murray et al. 2010). Further, high-quality longitudinal research has identified risk factors that increase the probability of problem behaviors in young people, such as violence, delinquency, teen pregnancy, substance misuse, and dropping out of school (Hawkins et al. 1992; Biglan et al. 2004). These factors are located in individuals but also families, peer groups, schools, and neighborhoods and communities. Protective factors, meanwhile, include skills to be successful, strong social bonds, opportunities for developmentally appropriate involvement, clear and consistent standards for behavior, and recognition for effort, improvement, and achievement (Hawkins and Weis 1985). Table 94.1 summarizes the specific indicators subsumed within the key developmental outcome and measurement framework. Some indicators are salient only at specific stages of children’s health and development. Table 94.2 provides examples of risk and protective factors.

Table 94.1 Summary of key developmental outcomes

Table 94.2 Sample risk and protective factors

In addition to strong predictive validity, the key developmental outcomes and risk and protective factors within this Evidence2Success framework needed to be measurable by parsimonious, valid, and reliable parent or child self-reports. A suite of measurement instruments was compiled to be completed by parents of young children aged 0–8 or children and young people aged 9–18 in no more than 1 h or 45 min, respectively. Critically, the outcomes also needed to be malleable by evidence-based interventions. In other words, it was a prerequisite that at least one evidence-based intervention exist to address a given key developmental outcome (e.g., absence of anti-social behaviour) or risk factor (e.g., poor parenting skills) or strengthen a protective factor (e.g., opportunities and rewards for prosocial involvement in the community). This is important as the tools are designed for use in communities and public service systems to identify and prioritize those specific key developmental outcomes within a population most in need of attention and help shift financial resources toward the targeted implementation of evidence-based approaches to improve these outcomes.

The application of surveys of child well-being, measuring a broad range of key developmental outcomes, risk and protective factors to inform strategic development and planning of services, is relatively rare. The Evidence2Success approach is currently being implemented in sites in the USA and Scotland. Perhaps, the most widely used approach to date, particularly in the USA, is Communities that Care (Arthur et al. 2007; Fagan et al. 2008). Data on child well-being are collected from representative populations of school children and used by community coalition boards to plan service delivery in the local area. The Social Research Unit has developed a similar approach in the UK for parents of children aged 0–8 and school children aged 9–18 for use by decision-makers in large public agencies (Axford and Hobbs 2010; Hobbs et al. 2010).

Key developmental outcomes and risk and protective factors essentially cover objective well-being. What about subjective well-being? Until recently, data on children’s subjective well-being was relatively scant (Bradshaw and Keung 2011a; Casas et al. 2012). Happiness and life satisfaction are commonly used as indicators of subjective well-being but they have limitations. Happiness implies short-term mood, but by focusing on the attainment of desires, it overlooks more functional aspects of well-being such as autonomy, self-acceptance, mastery, and personal growth (Searle 2008). Life satisfaction, meanwhile, is a more complex cognitive assessment of quality of life over time and an important predictor of subjective well-being (Searle 2008). But it shows a weak association with other important elements in determining subjective well-being, such as autonomy, personal growth, and purpose in life. There is a growing consensus that because subjective well-being is complex, it is best measured through obtaining information on a range of areas of experience, and not relying on single measures – hence, for example, the Oxford Happiness Questionnaire, which also covers ability to make decisions, feeling satisfied, having a sense of life achievement, and being in control (Hills and Argyle 2002). (The measure is not explicitly designed for children, although older children could use it.)

3.4 Measurement Issues

Defining well-being indicators as solely objective or subjective can be misleading. Most well-being constructs sit along a continuum of varying degrees of subjectivity. Few well-being indicators are truly objective in the way that mortality, income, and crime are. (Even these, which are administrative indicators, suffer from their own biases and error.) A global measure of happiness will usually be more subjective than a symptomatic measure of mental health, and measures of poverty or physical impairment will generally be more objective again. The level of subjectivity within a well-being indicator is bound up in the dimension of well-being selected, how the construct is being assessed, and if the measure is global (e.g., physical health) or more specific (e.g., a particular type of physical impairment).

“Subjective” infers personal feelings, opinions, and relativity. Thus, when an individual responds to questions about their physical and mental health, social networks, or socioeconomic status, their responses will be influenced by their belief system, knowledge, attitudes, emotional reactions, and cognitive appraisal. Take, for example, the Strengths and Difficulties Questionnaire (SDQ) (Goodman 1997), which essentially counts symptomatic behaviors relating to children’s mental health. It is likely that the meaning respondents attribute to these behaviors is affected by individual, social, and cultural factors. For example, one child might regard playing alone as lonely, whereas another child might regard this as desirable. Evidence on the equivalence of measures between different social groups is patchy, but one study showed that the SDQ performed differently between native and immigrant populations in Norway, concluding that this was likely due to the varying interpretation of questions and response options (Richter et al. 2011). Similar differences in measure performance have also found between countries when using purely subjective measures, such as the Personal Wellbeing Index (Casas et al. 2012).

This demonstrates that measures of behavior or mental health symptomatology, which tend to be thought of as relatively objective because they have been psychometrically validated, actually measure constructs that have subjective components. Poverty, which is typically assessed using measures of income or material deprivation, is also widely cited as an objective indicator of socioeconomic well-being. Yet research into measurement error in reported income shows that the level of misreporting and its direction (up or down the income scale) is explained by variables such as an individual’s work satisfaction, expectations, promotions they have received, and income itself (Akee 2007; Romanov and Gubman 2012). Moreover, measures of material deprivation sometimes overtly account for the subjective perspective. The Family Resources Survey (McKay and Collard 2004), for instance, includes items that were deemed essential by the public for living in modern British Society (Gordon et al. 2000; Pantazis et al. 2006) and in addition to asking respondents if they have or could afford each item it gives them the option to say that they do not need the item.

How objective a well-being indicator is may, in part, relate to an assessment of its psychometric performance, namely, its reliability and validity. In this context, reliability and validity can be viewed as an indication of objectivity. Reliability concerns how consistent a measure is; it does not account for the measure being accurate. For example, a ruler that is too short to measure a person’s height would reliably give the same measurement with each use, but this measurement would be inaccurate, as it would constantly underestimate height. Reliability is indicative of objectivity as one can assume when measuring a well-being outcome there is “noise” or error around the measurement and that this error will influence its precision or consistency. Some of this error will be random in nature; however, part of the error is likely to represent subjective influences (such as transitory moods and emotions).

It follows that well-being measures that are less objective may have lower reliability, as there is more potential subjective error to decrease the consistency of the measure. There is some evidence to support this claim (Kueger and Schkade 2008). For instance, test-retest reliability for measures of global satisfaction tends to be poor and indicates that such measures are quite subjective. At the other end of the scale, measures of demographic characteristics, such as income or educational level, have high test-retest reliability and may therefore be considered more objective. Measures of positive or negative affect sit somewhere in the middle.

Reliability is a prerequisite for validity. Validity captures how accurate a measure is, in other words how good it is at assessing the construct it claims to be measuring. Increasing subjectivity means a potential broadening of influences on the construct, and these influences represent yet more “noise” around the measure. This error will be correlated with a range of observed and unobserved characteristics of the individual (Bertrand and Mullainathan 2001). Increases in measurement error are a particular problem when looking at associations between well-being indicators. When subjective well-being is used to predict another indicator, measurement error is likely to undermine the association as it biases an association toward zero. This is due to the additional influence of unobservable personal characteristics (e.g., mood, cognitive appraisal). Of course, observable characteristics must be effectively controlled for otherwise the association will be further confounded. The impact of measurement error is more salient when subjective well-being is used as an outcome. In this case it is likely that predictor variables will correlate with the error in the subjective well-being measure, in addition to any correlation due to a true association (Bertrand and Mullainathan 2001). For example, suppose that individuals with a higher income have higher levels of life satisfaction (Ferrer-i-Carbonell 2005), but that higher income is associated with error in reporting of satisfaction: subjects with higher levels of income overreport high levels of satisfaction as this fits with their past behavior of earning more. In this case, it would be very difficult to disentangle the “true” effect of income on life satisfaction from the effect of income on life satisfaction measurement error.

All this said, subjective well-being indicators have been shown to be reliable and valid and are objective enough to be associated with other outcomes, risk, and protective factors (Trzcinski and Holst 2008; Van der As et al. 2010; Suldo et al. 2011). A review of predictors of subjective well-being (Dolan et al. 2008, p. 111) concluded that subjective well-being is “picking up differences in objective circumstances where we would expect to find them.” It also noted the difficulty in interpreting such associations due to the need to control for a large number of other influences and the role of reverse causality. Studies of income and subjective well-being, for example, suggest that higher well-being leads to higher income (Graham et al. 2004) but also that higher income leads to higher subjective well-being (Ferrer-i-Carbonell 2005).

To sum up, the vast majority of well-being indicators will include a level of subjectivity, and few are “objective” in a pure sense. Moreover, the level of subjectivity has implications for the statistical performance of a measure, and results need to be interpreted in light of this. This is not to say that highly subjective well-being indicators are not useful research tools, indeed there is an emerging consensus regarding their utility for predicting life course outcomes (which in turn influence on subjective well-being). While a hard distinction between subjective and objective well-being can be misleading, it is nevertheless helpful for broadly defining and measuring the concept of child well-being.

4 Connections and Contradictions

It is commonly assumed that objective and subjective well-being are aligned. Often this may be the case, but not necessarily. Children with good objective well-being may have poor subjective well-being, just as children with poorer objective well-being may have better subjective well-being. There are various possible reasons for this, as has been alluded to already. For example, subjective well-being is affected by prior experience and future expectations, which shape how current objective well-being is perceived. Similarly, a false awareness may mean that a child is not aware of what their well-being could be in a different context. Personal resources and traits also play a role: some people see the glass as half full, others think it half empty. As Searle (2008) put it, “It is individuals’ assessments of their circumstances, as opposed to their actual circumstances, that appear to have the greatest influence on reported well-being.”

Starting with the alignment of objective and subjective well-being, it is worth reiterating the distinction between different approaches to defining well-being. The hedonic approach emphasizes feelings of pleasure or happiness, whereas the eudaimonic perspective emphasizes modes of thought and behavior that provide engagement and fulfillment. Perhaps, unsurprisingly there is strong evidence of a close link between the two: “Overall, the available data indicate that the causal pathway between hedonic and eudaimonic well-being is bidirectional. Positive feelings can produce positive functionings and positive functionings can produce positive feelings” (Huppert 2005, p. 321). Another way of putting this is to say that while every individual has different preferences, there are many areas in which their utility or happiness are likely to be consonant with their interests – the need for food and shelter, the freedom to form friendships, the capacity to develop talents, and so on: “Something that is in a person’s interest can hardly fail to influence his state of mind. What a person does, what he achieves – more generally, what he is capable of achieving – are things that, while plainly not the sole determinant of his welfare [utility], are clearly one set of ingredients of his welfare [utility]” (Dasgupta 1993, p. 5).

Need fulfillment can be taken as a proxy of objective well-being. A recent study involving 123 countries examined the association between the fulfillment of needs and subjective well-being. Need fulfillment covered areas such as food, shelter, safety, security, and autonomy, while subjective well-being was measured in terms of life evaluation, positive feelings, and negative feelings (Tay and Diener 2011). The study found that need fulfillment was consistently associated with subjective well-being across world regions. Material circumstances can also be taken as a proxy indicator of objective well-being, and it is clear that these can shape subjective well-being. For instance, poor people often live in poor environments, producing a correlation between low income on the one hand and poor services, a lack of community spirit, and feeling unsafe in the neighborhood on the other (Bradshaw and Williams 2000; ONS 2000). Poverty also contributes to poor mental health – for instance, the constant need to juggle resources causes stress (Pantazis and Gordon 1997; Payne 2006) – and constrains a person’s autonomy, so that the poor must put disproportionate effort into survival (Bramley 1997).

However, individuals with poor objective well-being can still experience good subjective well-being. National and international reports show that most people report being generally happy or satisfied (Searle 2008), even when abused or unemployed (Diener 1993), and Ryff and Keyes (1995, p. 725) argue that people who live “ugly, unjust, or pointless lives” may experience feelings of happiness. Searle’s (2008, p. 52) analysis of data from the British Household Panel Survey (BHPS) during the 1990s shows that up to a third of people in the lowest social class (V) had some of the highest well-being levels.

Other examples reinforce the point. Chronically ill children can be surprisingly content (Graham et al. 1997). Or take the consumption of drugs and substances. This can lead to transitory and artificially induced feelings of “flow,” but in the long term, this and related behaviors “do not support refinement of any ability or competence, leading instead to health impairment and psychological maladjustment” (Delle Fave and Massimini 2005, p. 383). Similarly, some antisocial behaviors like stealing may be associated with subjective experiences of thrill, concentration, and high engagement (Delle Fave and Massimini 2005).

It is also possible for an individual to have limited material resources – again, a proxy for objective well-being – but good subjective well-being: the low-income episode might be brief, shallow, or offset by alternative resources such as savings (Ringen 1987). In addition, outlook, character, and constitution affect how much utility people derive from particular goods or circumstances (Sen 1982; Cohen 1993; Roker 1998). For example, people may be satisfied with very little if they have an ascetic bent or are ignorant of their relative socioeconomic status (Runciman 1966). A person may also choose to go without, indeed a small minority in developed countries adopt simpler, less stressful and more eco-friendly lifestyles: they “drop out” (Bliss 1996). There is a danger of romanticizing reality, but certainly there is evidence, partly from poor countries, that happiness is largely a product of things that money cannot buy – health, fulfilling relationships, natural beauty, and so on (Norberg-Hodge 1996) – and that hardship fosters solidarity: the phrase “slums of hope” has even been used to describe parts of some third world cities (Gordon and Spicker 1999).

The other side of the coin is that good objective well-being may be accompanied by poor subjective well-being. For example, paid work enables people to meet their needs, but for many whose employment involves laborious tasks, long hours, or unpleasant conditions, it is soul-destroying and requires sacrificing time with loved ones (Parker 1990; Leach 1994; Schluter and Lee 1993, 2002). Something that satisfies a persons’ need, and so contributes to good objective well-being, is unlikely to enhance a person’s subjective well-being if it is not of their choosing or taste. Further, what is sufficient to facilitate agency – in the sense of meeting basic needs (Doyal and Gough 1991) – is insufficient to help people attain the sublime, be that a personal ambition, say, a safari or playing a musical instrument well, or a sense of meaning (Doyal and Gough 1991, p. 189; Soper 1993; Miller 1999). Moreover, some people lack the necessary personal qualities to convert good objective circumstances into subjective well-being (Lane 1996).

Further, the analysis of BHPS data described earlier shows that those with the highest socio-economic status had some of the lowest levels of well-being during the 1990s (Searle 2008). One explanation for this is that the process of securing adequate resources often entails sacrifice and struggle – one reason why many developed countries have high rates of dislocated relationships and alcoholism (Argyle 2000). A second reason is that the happiness that people derive from improved living standards fades as they become accustomed to their new situation, or as they feel heightened security concerns, and also because they compare themselves against a new adjacent social group (to “keep up with the Jones’s”) (Layard 2003; Searle 2008). For example, although East Germans’ living standards soared after reunification in 1990, their reported happiness declined because they started to compare themselves more with West Germans rather than with Soviet bloc countries. Similarly, a study of 5,000 British employees found that reported satisfaction was more strongly related to relative income – how workers thought their income compared with that of similar colleagues – than to absolute income (Clark and Oswald 1995). Third, there are many important determinants of well-being besides socioeconomic status, and being economically secure is not protection in and of itself: thus, even the rich can be afflicted by, say, chronic illness or family discord. Taken together these points help to explain why there appears to be an income threshold above which extra income has little effect on happiness.

These patterns can be illustrated using data from a sample of over 8,000 children aged between 11 and 16 years from a Scottish local authority (Social Research Unit 2012). The Warwick-Edinburgh Mental Well-Being Scale is used here as a proxy measure for subjective well-being (Tennant et al. 2007). It comprises 14 positively worded items covering thoughts and feelings such as “I’ve been feeling cheerful” and “I’ve been feeling good about myself” and asks respondents to answer for the last 2 weeks using a five-point scale ranging from “None of the time” to “All of the time.” Three proxy measures for objective well-being were used (the study contained more). They are the “total difficulties” score on the Strengths and Difficulties Questionnaire (SDQ) (Goodman 1997), which indicates overall problems with behavior and mental health; the proportion of children eligible for free school meals (FSM), which indicates poor socioeconomic status; and physical health as estimated using the Kidscreen-52 (Ravens-Sieberer et al. 2001), which includes items such as “Have you felt fit and well?” (responses from “Not at all” to “Extremely”) and “In general, how would you say your health is?” (responses from “Excellent” to “Poor”).

The analysis shows that adolescents’ subjective well-being is associated with a range of objective indicators. Starting with the SDQ, a unit increase in total difficulty scores results in an effective difference of −0.44 in subjective well-being (95 % confidence interval, −0.46 to −0.42). When the SDQ is used to identify those with a likely behavioral or mental health disorder (the SDQ provides a cutoff score to indicate this threshold), there is an effect size difference of −0.86 (95 % confidence interval, −0.95 to −0.77) between this group and those with normal functioning. When children eligible for FSM are compared with the broader pupil population, there is an effect size difference of −0.24 for subjective well-being (95 % confidence interval, −0.31 to −0.17). Using Kidscreen-52, each unit increase in perceived physical health is associated with a 0.1 effect size increase in positive mental well-being (95 % confidence interval, 0.094 to 0.106), while children who perceive themselves as having significantly poor health show a −0.82 effect size difference for subjective well-being when compared with the broader population (95 % confidence interval, −0.88 to −0.75).

These patterns are as expected: they show that objective well-being is associated with subjective well-being. However, there is much variation between individuals. For example, 6 % of children eligible for FSM have very positive subjective well-being as measured using the WEMWBS (compared with 6 % of children not receiving FSM). Meanwhile, 6 % of children who do not receive FSM nevertheless have significantly negative mental well-being (compared with 8 % of those receiving FSM). Similarly, 3 % of children with normal psychopathology according to the SDQ report significantly negative mental well-being, compared with 21 % of children with mental health problems. And of children with a likely clinical mental health problem, 3 % report very positive subjective well-being (compared with 6 % with normal psychopathology). A small proportion of children (less than 1 %) have poor objective well-being across all three of the three objective well-being indicators: poverty, mental health, and physical health. Yet two in three children in this subgroup have typical subjective mental well-being. Finally, of the 773 children meeting the high need threshold of the SDQ total difficulties, just over half (52 %) think that the difficulties do not significantly impact on their lives. (The SDQ Impact Supplement assesses the impact of problems across the following categories: home life, friendships, classroom learning, and leisure activities) This highlights how salient perception is when assessing global well-being; that is, the construct is affected by one’s personal appraisal of objective conditions.

5 From Definition and Measurement to Policy and Practice

The users of data on objective and subjective well-being data, in particular directors or commissioners of services, need it to fulfill the following seven functions. The first four of these relate primarily to data on objective well-being.

5.1 Using Data on Objective Well-Being

The first function is epidemiological and involves counting the prevalence and distribution of disease, impairments, or other outcomes within a population. This principle may be applied to other fields – for example, counting the numbers of children living in poverty, or being maltreated, or displaying behavior problems, or misusing drugs or alcohol. In other words, the aim is to chart the pattern of objective well-being in a given population, be that a community, a country, or an administratively defined subgroup, such as children served by a social services agency. This type of descriptive epidemiology uses large data sets to describe something about a population – specifically what outcomes are found in whom at what time (Porta 2008). These methods can be used to describe patterns of well-being and so provide an empirical grounding for decisions about service provision (Axford and Hobbs 2010; Hobbs et al. 2010).

The second function concerns prioritizing which aspect of well-being, or which outcomes, to focus on. Once there is robust data on the prevalence of various outcomes and risk and protective factors, it is necessary to select those that might be the target of intervention or important to focus on in a context of limited resources. As indicated already, there are many dimensions of children’s health and development that services may seek to improve and many potential influences on children’s health and development – risk or protective factors – that services may seek to change. It is neither possible nor desirable to address all of these. The use of robust child well-being data, alongside appropriate banks of comparison or norm data, facilitates the process of prioritizing outcomes and associated risk and protective factors. Data may indicate outcomes where children are developing on a par with or better than children in comparable populations, or they may highlight deficits on which to focus efforts to improve outcomes.

For example, Fig. 94.1 shows those areas in which the average child in a large UK local authority is doing relatively well or poorly compared to national norm data (Hobbs et al. 2010). Bars reaching upward indicate that the average child in the local authority is doing better than the average child nationally, while bars falling below the central line indicate that children are doing worse. The magnitude of these positive or negative relative differences is indicated by effect size differences from the norm; those in the region of .2 are typically considered small but not necessarily insignificant, with those around .5 considered moderate and certainly significant (McCartney and Rosenthal 2000). In this example, children’s physical health and emotional health are relatively good compared with the norm, while hyperactivity and behavior present a mixed or broadly comparable picture. Both younger and older children’s social functioning is a cause for concern, however, suggesting that this might be an area of children’s health and development to prioritize for improvement.

Fig. 94.1

Prioritising outcomes: effect size differences from the norm

Third, objective well-being data can be used to estimate likely or realistic magnitude of improvement. Once the prevalence of outcomes and risk and protective factors is known, it is possible to plan and estimate the degree to which this may be changed by intervention. For example, the data in Fig. 94.2 illustrate that the behavior of the average child in the local authority referred to above is significantly worse than that of the average child in Britain (an effect size of − .38) and that nearly twice as many children have a likely conduct disorder (Hobbs et al. 2010). However, evidence-based programs for tackling poor behavior are unlikely to produce improvement large enough to equal British norms. A more realistic plan for improvement needs to be worked out at the population level, achieved by a balanced portfolio of targeted and universal evidence-based programs.

Fig. 94.2

Considering the mean and the tail

Fourth, all of the above may inform decisions about investment – whether in further research or in evidence-based programs, policies, practices, or processes. (A program is a discrete, organized package of practices, spelled out in guidance – sometimes called a manual or protocol – that explains what should be delivered to whom, when, where, and how. Policies refer to a course of action (or inaction) decided by policy makers to shape how people behave – for example, banning smoking in public places or withdrawing welfare to encourage people to find work. Practices refer to the activities of practitioners and may be broken down into discrete elements or techniques aimed at changing people’s behavior – for example, time-out, verbal praise, rewards, traffic light signals, tutoring, monetary fines, and mentoring (e.g., Mitchell 2011; Barth et al. 2012). Processes operate in service systems and include how youth and families are expected to access services, how their needs are assessed, and how staff are trained.) In the local authority cited above, data were used on both the well-being of the average child in a population and the proportion of that population in the tail of the distribution with likely impairments to their health and development as a device to inform children’s services directors’ and commissioners’ decisions about the type of interventions to be implemented to improve prioritized outcomes (Axford et al. 2012a). This was part of a £42 million investment in evidence-based prevention and early intervention (Little et al. 2012).

The data may indicate a large tail of the distribution comprising children with likely impairments to their health and development. Such children may benefit from targeted treatment interventions. For example, Fig. 94.2 shows that 21 % of children aged 11–15 in the local authority had likely conduct problems according to the SDQ, nearly twice as many as for the same age group nationally. Alternatively, the data may pave the way for a public heath approach in which universal services are used in an attempt to shift the whole distribution and so reduce the proportion of children with high-level needs (Rose 2002; Huppert 2005). Of course, these approaches are not mutually exclusive, and the data may help directors and service commissioners to consider the merits of both approaches in conjunction.

Objective child well-being data can also be used to help guide the selection of specific evidence-based programs – or other types of intervention – that have been demonstrated to improve that particular outcome, reduce risk, or enhance protection. In the local authority specified, a multiagency team of directors decided to focus on six outcomes (BCC 2007). These were selected because it was considered that they would enable most change and because they were related: improving one would improve others. A developmental approach was taken, meaning that the outcomes were mapped onto the different stages of children’s lives. Targets for children in early childhood, for example, were different from targets for adolescents. A distinction was also made between all children and children in need. Again, the targets for these two groups of children were different. Taking emotional health as an example, the aim was to make all children a bit happier but also to reduce the number of children with depression.

The team of directors then identified activities geared toward improving the priority outcomes. These included universal services for all children and targeted early intervention for children at risk. PATHS (Promoting Alternative Thinking Strategies) was selected under the former. It is a school-wide social and emotional learning curriculum for children aged 4–11 years and aims to develop children’s self-awareness, motivation, empathy, social skills, and ability to manage feelings (Greenberg and Kusché 2002). It was selected to help address identified deficits in prosocial behavior and peer relations (Fig. 94.1). The Incredible Years BASIC program was chosen as a targeted early intervention program. It is a 14-week group-based parenting course for parents of children aged 3–4 with conduct problems (Webster-Stratton 2007) and was selected to reduce the prevalence of conduct disorder in later years. Both programs have been found to be effective in multiple randomized controlled trials and are implemented in several countries.

5.2 Using Data on Subjective Well-Being

Promoting children’s subjective well-being seems a reasonable goal for any society (Bradshaw and Keung 2011a). But there is a live debate about the extent to which subjective well-being is “policy salient.” This centers on two questions: can average levels of children’s subjective well-being really improve in the longer term, and can policy really influence this? Evidence that personality traits and genetics (Diener et al. 1999) or psychological devices that protect a normally positive mood (Cummins 2009) are important factors in determining subjective well-being has tended to lead to skepticism on this issue. However, data showing that children’s average level of happiness amongst 11–15-year-olds in the UK improved over the period 1994–2008 offers “a tentative challenge to that view” (Bradshaw and Keung 2011b, p. 15). (This is based on an analysis of the British Household Panel Survey, a cross-sectional survey conducted annually and from 1994 administered to 11–15 s as the “British Youth Panel Survey.” It asks respondents to score how they feel about their school work, appearance, family, friends, and life as a whole on a six-point Likert scale, where 0 = not happy at all and 6 = very happy (yielding a total happiness score of 30)). Why it improved is less clear, but there is a suggestion that children’s relationships with friends and their happiness with school – both of which may be amenable by policy – played a part.

Policy efforts tend to focus on objective circumstances, for the reason alluded to earlier in this chapter: while every child has different preferences, there are many areas in which their utility or happiness are likely to be consonant with their interests. However, some aspects of children’s services might be interpreted as addressing subjective well-being. Thus, the concern with eliciting and heeding individuals’ preferences plays out in the drive toward joint decision-making; for example, family group conferences are intended to empower families with problems to take control of their situation and formulate solutions, although evidence for their impact is scant (Sundell and Vinnerljung 2004; Berzin 2006). Also relevant are efforts to improve children’s personal resources so that they gain greater benefit from their objective circumstances (Lane 1996). In particular, education is critical in enhancing children’s cognitive complexity and in training them to appreciate beauty or exposing them to spiritual and other uplifting experiences. Much moral and religious teaching in school seeks to foster a questioning attitude toward western materialism and the individualist ethos; it encourages pupils to be content with less and to value friendships and tolerance. There has also been a concerted effort in recent years to promote more emphasis in schools on the social emotional, creative, and physical aspects of development, as well as on academic achievement (Huppert 2005; Layard 2005).

Some of the policy recommendations in this area focus on increasing what psychologists refer to as “optimal experience” (or “flow”). This results when a person is fully engaged in a task in which his/her interests are fully utilized (Csikszentmihalyi 1997). It is characterized by “concentration, focused attention, positive mood, clear goals, unselfconsciousness, and intrinsic motivation… [P]eople describe themselves as […] excited and relaxed at the same time” (Delle Fave and Massimini 2005, p. 382). Or, as Huppert (2005, p. 319) put it, “Optimal experience arises when the person’s level of skill matches their level of challenge; subjective well-being is compromised if the challenge is either too difficult, leading to frustration, or too easy, leading to boredom or apathy.” A consequence of this is that while the way in which children respond to their environment affects their subjective well-being, they can also play an active role in shaping their subjective well-being. This is through what Sheldon and Lyubomirsky (2004) call “intentional activities.” Those related to subjective well-being include: overt behaviors, such as exercising or being kind to others; cognitions, such as interpreting events positively or enjoying the moment; and motivations, such as striving toward valued – and, critically, self-generated – goals or trying hard at worthwhile activities (Huppert 2005). Policy and practice could therefore improve subjective well-being by seeking to modify behaviors, cognitions, and motivation.

A fifth function of well-being data therefore involves seeking to understand what makes children and young people happy and fulfilled. This can be used to understand subjective well-being, regardless of objective indicators. It may also be used to assess the impact of objective indicators of influences upon subjective well-being. In the UK, The Children’s Society has explored this question in depth. They have found from repeated national surveys that, at any given time, around 14 % of 15-year-olds have low “subjective well-being” (Rees et al. 2012). A number of key themes have emerged from this work.

While the research has indicated that relationships with friends and peers are important, children’s feelings about the quality of relationships with key adults in their lives appear to be even more important: the amount of care and support children receive from parents, teachers, and other key adults is a key theme, as is adults treating them with respect, recognizing their increasing competencies as they grow up and allowing them freedom of choice and expression. Where children have poor relationships with adults in the immediate family, the value of good relationships with other adults becomes all the more important (Rees et al. 2012). Other research highlights importance of relationships for subjective well-being (Searle 2008); indeed, strengths that enable individuals to build relationships with others are more predictive of subjective well-being then self-predicted future well-being (Gillham et al. 2011).

A healthy “work-life balance” is acknowledged as important for adults’ general well-being. The same appears true for children. Many children in the study by The Children’s Society mentioned academic work pressures as something that prevented them having a good life. Generally speaking, children who felt they spent too little time with family and friends had lower overall subjective well-being (Rees et al. 2012). Unsurprisingly, money and possessions are also important, but “having enough” is much more important to children than being wealthy (Rees et al. 2012). Those who felt they had roughly the same amount of money as their friends had the highest overall levels of well-being. Thus, while living in poverty is a predictor of poor subjective well-being, relative deprivation appears a much stronger driver in general terms. For older young people, another key factor contributing to subjective well-being is likely to be employment (Searle 2008). This yields several benefits, including social interaction. Social interaction provides a source of identity, feelings of belonging, and a sense of purpose (Jordan 2008) and also buffers individuals from stress (Veenhoven 1984). Indeed, such is its importance that Jordan (2008) suggests that relationships should form the basis of social policy.

5.3 Using Objective and Subjective Child Well-Being Data

A sixth function of well-being data is to understand association and causality. If data on subjective or objective well-being are collected in combination with other sources of data, it is possible to examine associations between them (in the case of cross-sectional data) or, if the data are collected longitudinally or manipulated experimentally, to examine causality. It is valuable to measure the association between outcomes and known or hypothesized risk and protective factors (Porta 2008) because the distribution of well-being and risk and protective factors – and therefore the strength of associations between them – varies by population (Bradshaw et al. 2009; Annie E. Casey Foundation 2011). This is known as analytical epidemiology (Porta 2008) and concerns how and why outcomes are found in particular individuals at certain times. Evidence-based programs typically work by addressing risk and protective factors through a theory of change, so the right interventions are most likely to be commissioned when both the distribution of well-being and the likely reasons for its distribution are known.

To illustrate, take the finding that children in a given jurisdiction have elevated levels of conduct disorder. How should services respond? Impulsiveness and violent interparental conflict are two of many important predictors of conduct disorder (Murray and Farrington 2010), so an analysis of whether and by how much either risk factor contributes to behavioral problems in this specific population would help guide planners on whether they should prioritize addressing impulsiveness or violent interparental conflict. The Communities that Care survey, for instance, which has been implemented fairly widely in the USA and Europe, allows this to be done (Hawkins et al. 1992; Arthur et al. 2007; Fagan et al. 2008).

Seventh, both objective and subjective well-being indicators may be measured over time, and the impact of context on these indicators may be examined. This enables researchers and policy makers to monitor and understand trends – whether aspects of child well-being are remaining stable, deteriorating, or improving. Without a control group, it is difficult to attribute changes to a specific policy or intervention because of possible confounding factors, such as changes in the wider context. Nonetheless, these indicators cautiously indicate the direction of change and can be used the “take the pulse” of child well-being in a given context (Bradshaw et al. 2010, 2011; Whiteley et al. 2010). In Birmingham, the survey referred to earlier has been implemented annually since 2007. Levels of child well-being have remained broadly stable over this five-year period. In the USA, a large multistate experimental trial of Communities that Care found population-level changes in risk and protective factors over a two- to three-year period, with changes in outcomes occurring at 4–5 years (Oesterle et al. 2010).

6 Future Directions

In this closing section, we outline ways in which the field needs to develop, starting with conceptual issues and then moving through measurement and analysis to the way in which data on children’s objective and subjective well-being can be used.

6.1 Concept

The concepts of key developmental outcomes and risk and protective factors as outlined earlier in this chapter deserve to be used more widely. These are based on extensive reviews of the literature and will help to bring more order and consistency to the way in which objective child well-being is measured.

Meanwhile, greater care is needed in the way in which subjective well-being is defined and operationalized. Specifically, there are discrete types of subjective well-being, notably life evaluation, positive feelings, and negative feelings (Lucas et al. 1996). Studies need to be explicit about which of these they are focusing on. This is particularly important when it comes to exploring associations between objective and subjective well-being. Taking need fulfillment as a proxy of objective well-being, recent research suggests that the fulfillment of some needs is more strongly associated with some types of subjective well-being than others. For example, Tay and Diener (2011) found that life evaluation was most strongly associated with the fulfillment of basic food and shelter needs, whereas positive feelings were more associated with the fulfillment of needs for social support and feeling respected.

6.2 Measurement

The way in which aspects of objective and subjective well-being are measured and reported needs to be tightened up. At present, much caution is required when making comparisons between studies assessing well-being. When global measures of well-being are used, there is variation in the constructs used. Greater clarity can be achieved by using case definition. This approach draws from medical epidemiology and involves specifying precisely the method of measurement: the constructs contributing to well-being, how they have been assessed, the unit of analysis, and the boundaries of a condition. For example, if a study aims to assess the effect of parental employment on children’s subjective well-being, the association will be affected by how the independent variable (employment) is categorized: does “unemployment” include parents who cannot work because they are disabled or looking after children full time? The contextual information included in a case definition facilitates correct interpretation and therefore comparison of results across studies.

6.3 Analysis

There is scope for much more research on the connections between objective and subjective well-being overall but also between their constituent parts. This should involve using simpler and more complex variations of subjective well-being – in other words, using happiness or life satisfaction but then also taking into account factors like autonomy and sense of mastery or separating these out even further. It is particularly interesting to understand more about why and how disjunctions arise, in other words when a child with good objective well-being has poor subjective well-being, and vice versa.

More care is also needed when exploring the association between risk and protective factors and aspects of objective and subjective well-being. Models seeking to explore variation in well-being would ideally indicate the relative contribution of individual predictors alone and in combination. Studies should also take greater account of group-level variables that might explain variation in well-being. Often, models focus on individual factors only and ignore the group level, such as households or schools. Relevant class level factors that may affect the well-being of pupils, for example, those who perform badly in an academic test, include overall class performance, class attitudes about academic achievement, and what the class expects pupils to achieve (Tamir et al. 1985; Cialdini et al. 1991). Similarly, a review of neighborhood determinants of children’s health and well-being found that, when individual and family characteristics were controlled for, on average 10 % of variance in health outcomes was explained by children’s neighborhood (Sellström and Bremberg 2006).

If research on child well-being is to have more influence on policy and practice, there is a need to explore in more depth the associations between objective and subjective well-being on the one hand and service use and outputs on the other. The purpose of this is to understand more about predictors of service use and the relationship between outcomes, which matter to individuals, and outputs, which matter to systems. The better measurement of service use is a necessary first step, since until now a lack of adequate data has made it difficult to know exactly what children in a given service system or geographical area actually receive. Drawing on research in child mental health, a new measure of service use covers critical service dimensions, including provider, setting, duration, intensity, and content, thereby enabling a stronger description of what children and families receive (Axford 2010). Meanwhile, a more sophisticated understanding of how changes in child objective and subjective well-being might affect outputs is critical because of the financial implications involved. For example, does reducing child anxiety and depression affect the number of children admitted to psychiatric care? Or does improving children’s behavior affect rates of entry into the youth justice system? Ultimately, this information should help children’s services directors and commissioners to invest in interventions known to improve outcomes because they will see the likely beneficial effects for the output indicators against which children’s services agencies are traditionally judged. The information will also help, potentially, to reallocate resources toward early intervention premised on anticipated future savings on more resource-intensive forms of intervention.

6.4 Use of Data

Methods are needed to enable service directors and commissioners to use data on the well-being of local children to prioritize areas for improvement and then select tested and effective interventions that will prevent or address the problem identified. These should include programs but also, as already noted, policies, practices, and processes. The Blueprints for Healthy Youth Development database includes evidence-based programs that meet standards agreed by a respected international panel of leading scientists (Axford et al. 2012b). Programs on the database have been evaluated by at least one good randomized controlled trial or two good quasi-experimental studies with consistent evidence of positive outcomes. They must also be ready for dissemination in communities and service systems. The database can be searched for programs that address outcomes, risk, and protective factors of interest. A well-being instrument that can be implemented simply in schools and the local community has been designed to link directly to this database. Thus, if data from the well-being instrument indicate a problem with drug misuse, users will be directed to a program such as Life Skills Training (Botvin and Griffin 2004), or if the identified difficulty concerns depression or anxiety, it would point users to, say, the Penn Resiliency Program (Brunwasser et al. 2009). There is information on the nature, value, and implementation requirements of each program, including outcomes targeted, intended recipients, costs, and cost-benefit ratio.

There is also a need to make data “live,” in other words to get it into the hands of directors, commissioners, managers, and practitioners rapidly and in a format that is accessible and easy to manipulate and analyze. Several so-called continuous feedback systems are being tested in the USA and can be used for several purposes (Bickman et al. 2011; Cash et al. 2012). Some of these relate to service delivery, including fidelity and quality of implementation and the potential need for additional training for staff who are struggling, but others concern monitoring changes in child well-being in a given service or area and determining if clients respond better to some treatments than others.

Finally, although there are numerous initiatives designed to improve children’s subjective well-being, the evidence base for them needs strengthening. There is a need to identify interventions that meet high standards of evidence, such as the standards used by Blueprints, but also to enable other interventions to reach those standards. Population-level public heath style interventions offer particular promise: examples include education or health promotion approaches through schools, workplaces, heath clinics, and community centers. There is also scope for the media to provide more reliable information and model desirable behavior: television soap operas, for example, tend to normalize drug misuse, eating disorders, and violence; they could portray and thereby promote tolerance, compassion, and responsibility (Huppert 2005). But such approaches need to be evaluated rigorously in quasi-experimental or randomized controlled trial studies. There are already several reviews of successful school-based programs aimed at preventing child behavioral and mental health problems (e.g., Greenberg et al. 1999; Weissberg et al. 2003; Durlak et al. 2011), and there is emerging evidence of the benefits of mindfulness interventions with children (Huppert 2005; Mendelson et al. 2010; Greenberg and Harris 2012). (Mindfulness is “the state of being attentive to and aware of what is taking place in the present” (Huppert 2005, p. 324).)

7 Summary

There have been concerns in recent years about deficits in children’s well-being despite increasing affluence in many societies. These concerns have led to a greater policy focus on child outcomes and in-depth discussions of what a “good life” means for children. The distinction between objective and subjective well-being is a major fault line in these debates. Objective well-being covers different dimensions of children’s health and development – behavior, emotions, attainment, and so on – and, to some extent, the factors that impinge on these, such as housing, parenting, environment, and socioeconomic situation. Subjective well-being concerns how children and young people assess their lives, in particular how satisfied or happy they are with aspects of their lives or their lives overall.

The way in which well-being is defined and measured is shaped by different ways of conceptualizing “the good life.” One view is that it involves maximizing happiness and satisfying people’s preferences; another focuses on the substantive goods that always and everywhere make life better. A third perspective is that neither approach in isolation captures well-being: both are needed. One argument for combining them is that objective and subjective well-being can be at odds for various reasons – for example, people adapt to new circumstances, and personal traits affect how much benefit people derive from their situation. A key message is that self-determination – the capacity to exercise choice in forming and pursuing a coherent life plan – is an important determinant of well-being.

The measurement of well-being owes much to social indicators movement (from the 1960s onward) and research on health-related quality of life (from the 1970s). There has been a shift in what is collected, for example, with the increasing use of indicators of positive well-being. Health has made good use of psychometrically validated scales, of which there are also now many in other areas of children’s objective well-being, and there has been an increase in measures of children’s subjective well-being – both in a global sense but also for sub-domains. Generally, there has been a concerted move toward using children as the primary source of data for many aspects of well-being on the basis that they are active participants, not passive subjects, and they know their situation and feelings better, often, than adults.

A recent development in measuring objective well-being is the use of key developmental outcomes – those aspects of children’s well-being that are most critical for children’s subsequent health and development. These fall into five broad areas: behavior, emotions, relationships, physical health, and educational skills and attainment. If children are not achieving the specified outcomes at particular stages of development, there is a strong likelihood that their future health and development will suffer. Regarding subjective well-being, the limitations of measuring single constructs like happiness are increasingly recognized. There is therefore a growing view that because subjective well-being is complex, it is best measured through obtaining information on a range of areas of experience and/or disaggregating the construct.

It is accepted that few measures of well-being are truly objective – even measures of behavior, material well-being, or health. Even so, it is important to strive for reliability and validity in measurement. There have been particular concerns in this respect with measures of subjective well-being, but there is an emerging consensus regarding their utility for predicting life course outcomes.

It is commonly assumed that objective and subjective well-being are aligned. This is often but not always the case. Children with good objective well-being may have poor subjective well-being, just as children with poorer objective well-being may have better subjective well-being. There are various reasons for this, but an important factor is that individuals’ assessments of their circumstances, as opposed to their actual circumstances, appear to have the greatest influence on reported well-being.

Data on objective and subjective child well-being need to be used to inform policy and practice. Objective well-being data can be used to chart the pattern of objective child well-being in a given population; prioritize which aspect of well-being to focus on; estimate likely or realistic magnitude of improvement; and inform decisions about investment in evidence-based programs, policies, practices, or processes. Data on subjective well-being is important from a policy perspective because subjective well-being can be changed – not just by improving objective well-being but also by affecting children’s cognitions, behaviors, and motivation. Objective and subjective child well-being data can be used to understand association and causality, for example, what causes problems or what makes children and young people happy and fulfilled. It can also be used to monitor trends over time and so indicate whether aspects of child well-being are remaining stable, deteriorating, or improving.

Work in the area of objective and subjective child well-being needs to advance in several directions. Conceptually, the notions of key developmental outcomes and risk and protective factors need to be used more widely and consistently, and greater clarity is needed about which aspect(s) of subjective well-being are being studied. There is an ongoing need to analyze the relationship between objective and subjective well-being, with particular attention to reasons for disjunctions between them. If research on child well-being is to have more influence on policy and practice, there is a need to explore more fully the associations between objective and subjective well-being on the one hand and service use and outputs on the other. Methods are needed to enable service directors and commissioners to use data on the well-being of local children to prioritize areas for improvement and then select tested and effective interventions that will prevent or address the problem identified. In this respect, the evidence base for interventions designed to improve children’s subjective well-being needs strengthening.