The Welfare Trait

by Adam Perkins

  The rational response in the UG is to accept any offer, however small, since something is always better than nothing from the standpoint of maximising economic gain. If this rule applied in humans, every UG offer, however mean/selfish, would be accepted. Yet scores of studies have shown that humans do not usually behave rationally in the UG. Most UG research to date has been in undergraduates from advanced nations: in this population, offers are on average between 40 and 45 per cent and offers of this size are usually accepted. Offers substantially below this value, such as 20 per cent, are typically rejected about half the time, as if to punish the proposer for being selfish (Sanfey et al., 2003). This deviation from rationality in the UG suggests that humans care about fairness and cooperation, but undergraduate students in advanced nations are not necessarily typical of the wider human population. This factor means that these early UG data were unable to reveal whether a fair-minded attitude towards non-family is universal across human cultures or whether it is shaped by local factors.

  In an attempt to resolve this problem, Joseph Henrich and colleagues administered the UG to the members of 15 primitive, small-scale, mainly tribal hunter-gatherer or forager societies. This research revealed significant group-level differences between the 15 societies in attitude towards individuals from outside the family unit (Henrich et al., 2005). Whether these differences in attitude are wholly learned or also have a genetic component is unclear, but given what we know about the biological basis of personality, they are likely to be a product of both nature and nurture acting in response to the environment. More specifically, the differences in attitude between the societies studied by Henrich and colleagues seem to be shaped by the degree to which survival in a particular environment demands group-level cooperation. This idea is supported by the finding that the societies whose members care little about fairness (as evidenced by low offers being made and accepted in the UG) tend to occur in environments in which cooperation with non-family members has negligible survival value. Conversely, the societies whose members typically care strongly about fairness (as evidenced by high offers being made and accepted in the UG) tended to occur in environments in which cooperation with non-family members has major survival value.

  As a caveat, these discoveries should not be interpreted as meaning that the members of small-scale societies all have the same personality profile, because there is evidence that personality outliers exist in primitive, hunter-gatherer societies, just as they do in advanced nations. For example, anthropological studies of Eskimo culture have revealed norms that are harsher than those in a typical advanced nation, as in times of hardship, the Eskimo traditionally conserved food by practising both infanticide and senicide (the killing of children and elders, respectively).

  Yet evidence exists of individuals who repeatedly breach the norms of Eskimo society and, furthermore, these individuals have approximately the same personality characteristics as an employment-resistant individual in a developed Western nation such as the UK or USA (that is, who combine low conscientiousness and low agreeableness). For example, the Yupik of Northwest Alaska label unconscientious and disagreeable individuals as kunlangeta and describe them as follows:

  The man who, for example, repeatedly lies and cheats and steals things and does not go hunting and, when the other men are out of the village, takes sexual advantage of many women – someone who does not pay attention to reprimands and who is always being brought to the elders for punishment. One Eskimo among the 499 on their island was called kunlangeta. When asked what would have happened to such a person traditionally, an Eskimo said that probably ‘somebody would have pushed him off the ice when nobody else was looking’.

  (Murphy, 1976, p. 1026)

  The UG research by Henrich and colleagues has identified two small-scale, hunter-gatherer societies that possess particularly extreme cultural norms connected to fairness and cooperation, namely the Machiguenga of the Peruvian Amazon and the people of the village of Lamalera of Indonesia. In the UG, the Machiguenga on average offered 26 per cent of the asset to the responder and the highest offer by any Machiguenga was 50 per cent. Similarly, only one out of ten Machiguenga responders rejected an offer of below 20 per cent (Henrich, 2000). Overall, this pattern of results shows that compared to undergraduates from advanced nations, the Machiguenga possess a selfish attitude towards non-relatives, with a minimal sense of fairness, and feel little or no obligation to offer an equal share to responders. In line with this attitude, the Machiguenga had low expectations of generosity from others and seemed not to harbour any desire to punish selfish or mean proposers.

  In contrast, the lowest Lamalera offer was 40 per cent of the asset, the average offer was 57 per cent and the highest offer was 90 per cent (Henrich et al., 2005). Since none of the Lamalera made a low offer, the experimenters introduced 20 sham offers ranging from 10 to 50 per cent of the asset (average 30 per cent) in order to assess the attitude of the Lamalera towards selfish proposers. Almost half of the lowest sham offers (three out of the eight offers) were rejected. Overall, this pattern of results suggests that the Lamalera are unselfish, with a strong sense of fairness. They are therefore closer in attitude to the undergraduates in advanced nations than they are to the Machiguenga, despite being hunter-gatherers like the Machiguenga.

  The most likely cause of these attitude differences between the Machiguenga and the Lamalera is the environmental differences between these two societies: the Machiguenga have for centuries pursued a nomadic hunter-gatherer lifestyle – bolstered by some slash and burn horticulture – in the jungles of the Amazon basin in south-eastern Peru. Due to this lifestyle, small (typically single-family) groups of Machiguenga can survive independently, without the need to cooperate beyond the family unit. Even in recent years, when the influence of missionaries and other outside agencies has caused the Machiguenga to settle in villages of approximately 300 people and turn to a largely horticulturalist existence, the Machiguenga remain almost completely economically independent at the family level (Henrich, 2000).

  The people of the village of Lamalera, on the south coast of Lembata Island in Indonesia, are also hunter-gatherers but with a very different environment to the Machiguenga, since Lembata Island is poorly suited for supporting crops or game animals. Instead, the Lamalera have historically survived by hunting large sea animals, especially sperm whales (Alvard, 2011). This is accomplished using primitive technology, namely the village’s flotilla of approximately 20 wooden rowing boats (known as téna), each typically 11 metres long and crewed by approximately 15 men who are equipped with hand-thrown harpoons. The limitations of their basic equipment mean that a successful whale hunt requires that the Lamalera behave in a highly collectivised and organised manner that cuts across family units: when whales are sighted, the village mobilises as a team in order to attempt a catch. The tightly choreographed whale hunt is directed by the boat captains, known as lama fa, who tend to be the most experienced and skilful harpooners.

  If a whale hunt is successful, the resulting meat is shared amongst the households of the village irrespective of family relationships and including elderly and sick residents who have played no part in the hunt. Meat is also traded with other non-whaling communities nearby. The difficulty of killing and retrieving whales with primitive equipment means that only a few per cent of whale hunts are successful: despite hunting whales at near weekly intervals from May to October, the Lamalera typically catch fewer than six whales a year.

  These UG studies are important to the argument in this book because they suggest that, if survival can be accomplished without cooperation beyond the immediate family, humans tend to gravitate towards a selfish, disagreeable attitude, as demonstrated by the Machiguenga. Since the welfare state permits survival without cooperation, it therefore risks encouraging claimants to develop a pseudo-Machiguenga attitude – or, as I have already dubbed it, the employment-resistant personality profile. Furthermore, because the asset used in the UG is provided for free by the experi
menter, these data suggest that there is a causal link between personality and a willingness to accept handouts, because the UG results suggest that uncooperative individuals (that is, individuals with low scores on agreeableness) are especially greedy when it comes to accepting free resources. These UG data therefore bolster my argument in Chapter 2 that the welfare state filters employment-resistant individuals into the benefit-claiming sector of the population because individuals with low levels of agreeableness (the Machiguenga) are especially susceptible to the lure of unearned income. In line with this idea, studies have shown that antisocial personality characteristics are over-represented amongst welfare claimants (Vaughn et al., 2010).

  A sceptic might argue that the agreeable, fair-minded attitude of the Lamalera is not a genuine example of non-family cooperation, but is actually a form of nepotism that evolved by natural selection due to close genetic relatedness between the residents of Lamalera. Due to the village’s isolated location, this is plausible: for example, even though the UG game is played anonymously, the Lamalera participants in the UG might have assumed that the other player must be a relative because everyone in the village is closely related. In line with this notion, altruism between close relatives such as parents and siblings (that is, nepotism) can evolve by natural selection when the donor and recipient share many genes, as this means that the donor is effectively helping themselves (Hamilton, 1964).

  But the cooperative behaviour of the residents of Lamalera cannot be explained away as nepotism because in hunter-gatherer terms, Lamalera is a relatively large settlement, which contained approximately 1,200 people at the time of the UG research. The population of Lamalera is therefore too large to be viewed as a single-family unit equivalent to a traditional Machinguenga grouping. In line with this notion, research has confirmed that close genetic kinship plays little part in the whale-hunting organisation of Lamalera. For example, no set of siblings in the village was large enough to crew even a single téna (whale-hunting boat), let alone the entire flotilla of 20 téna (Alvard, 2011).

  The Lamalera whale hunters therefore provide a prime example of what is known as cultural kinship, in which a large group of unrelated individuals cooperate to accomplish a complex task (in this case, killing and retrieving a whale) that is impossible for a lone individual or a family group. Briefly stated, cultural kinship is the phenomenon that opens the door to the modern, organised world of altruistic human endeavour – of which the welfare state is, ironically, a prime example. This idea is backed up by the previously mentioned finding that modern, advanced populations (that is, undergraduates) perform similar to the Lamalera on the UG (Sanfey et al., 2003).

  Viewed as a whole, these UG studies suggest that a welfare state which provides handouts without requiring work in return will encourage the development of employment-resistant characteristics by weakening cultural kinship. This analysis echoes the concerns of Scandinavian economists that the welfare state erodes work motivation by weakening work-related norms (Lindbeck, 1995), but it does not contain a reproductive component. It therefore might seem to be of little relevance to the welfare trait theory which maintains that changes in work motivation are influenced by welfare-induced differences in the number of children born to claimants and workers.

  In order to make the relevant link, it is necessary to realise that these UG data dovetail with the concept of r–K selection which was introduced in Chapter 4. As we saw in that chapter, in conditions where resources are plentiful and competition for these resources is low, the optimal reproductive strategy is r selection, which entails rapid reproduction with relatively little regard to the wellbeing of the resulting offspring. Conversely, when resources are scarce and have to be competed for, the optimal reproductive strategy is K selection, which entails producing fewer offspring but nurturing them carefully so that each offspring is itself capable of competing for resources.

  In Chapter 4 we saw evidence that employment-resistant personality characteristics (low levels of conscientiousness and agreeableness) are associated with r selection whereas pro-employment personality characteristics (high levels of conscientiousness and agreeableness) are associated with K selection. Since the Machiguenga live in an environment in which resources are relatively plentiful (the Amazon basin) whereas the Lamalera live in an environment in which resources are relatively scarce (a small, stony island), we should expect the Machigeunga to incline towards r selection and the Lamalera towards K selection, with all the personality implications that accompany these two strategies.

  The UG data obtained by Henrich and colleagues are only able to shed light on the agreeableness-related aspects of this hypothesis, but they fit it well, since they suggest that the Machiguenga are indeed significantly less agreeable than the Lamalera. In the context of this book, the welfare state can be conceptualised as simulating the Machiguenga environment in that it removes the need to cooperate with non-family in order to survive and also removes the need to care for children conscientiously. If this idea – and thus the welfare trait theory – is valid, we should see evidence that both antisocial behaviour and large, neglected broods of children (two correlates of r selection) are especially common amongst welfare claimants. The next section presents evidence of this type, using data from the Troubled Families Programme.

  The Troubled Families Programme

  The Troubled Families Programme was initiated by the UK government in December 2010, with all 152 upper tier local authorities in England being asked to identify families in their catchment area who required extra help, due to meeting three out of the four following criteria:

  1. Are involved in youth crime or antisocial behaviour.

  2. Have children who are regularly truanting or not in school.

  3. Have an adult on out-of-work benefits.

  4. Cause high costs to the taxpayer.

  This initiative revealed that there were 111,574 families in England that met these criteria and could therefore be defined as ‘troubled’. The entry criteria used to select families were relatively basic and so, to capture more fully the characteristics of these families, the government then asked each of the 152 upper tier local authorities randomly to select ten per cent of the troubled families in their catchment areas. The local authorities were asked to provide information on the selected families concerning 35 important variables, including employment, education, crime, housing, child protection, parenting, family size and health. Information was received from 133 of the local authorities, covering 8,447 families. In 1,048 of these families, data were available on all 35 variables, providing a comprehensive and detailed picture of their characteristics. Analysis of the data relating to these families is contained in the report entitled ‘Understanding Troubled Families’ that was published by the Department for Communities and Local Government in July 2014.

  Overall, the report revealed that although the families had been selected on the basis of suffering three problems, on average, each troubled family actually suffered from nine problems. The unexpectedly severe magnitude of the difficulties suffered by troubled families prompted the authors to observe:

  For any family facing just one or two of the problems of the type highlighted above there is a higher risk of poor outcomes for their children. For example, truancy and parental unemployment tend to increase the likelihood of a child or young person becoming a NEET (that is a 16–24 year-old not in employment, education or training) with all the disadvantages this brings. However for troubled families, with an average of nine problems, the cumulative effect of these problems is likely to make it more difficult to get that child back into school, to tackle criminal behaviour or get a parent into work.

  (Understanding Troubled Families, p. 11)

  The report itemised the problems faced by trouble families, with two of the most prominent problems being that 83 per cent contained a workless adult and 54 per cent of the troubled families were involved with crime or antisocial behaviour. These findings are congruent with my hypothesi
s that the employment-resistant personality profile is just one manifestation of a general tendency to behave in an antisocial manner. However, since unemployment and antisocial behaviour constituted two out of the four entry criteria for the Troubled Families Programme, this association is unsurprising. Of more interest in the context of this book are the report’s additional findings concerning family size that were not part of the entry criteria to the Troubled Family Programme. If these characteristics cluster in a way that fits the welfare trait theory then they will provide support for it, since the entry criteria of the programme cannot be responsible for this pattern.

  In short, the additional findings of the report fit the welfare trait theory closely. For example, as I outlined in Chapter 1, the central postulate of my theory is that the welfare state proliferates the employment-resistant personality profile by boosting the number of children born into disadvantaged households. For this postulate to hold true, it should be the case that troubled families, since they are particularly rich with employment-resistant personalities, will have more children than typical workless families and also that the problematic personality characteristics of the adults that cause families to be troubled in the first place are transmitted to the children. Both these points are supported by the report.

  First, the report found that the troubled families on average contained 2.5 children. We have already seen in Chapter 4 that across the whole UK population there is a positive linear association between number of children and reliance on the welfare state: the greater the degree of reliance, the more children in the household (Figure 9.1).