The Welfare Trait

by Adam Perkins

  However, raw, census-style data such as those shown in Table 4.1 take no account of confounding factors. For example, having an employment-resistant personality profile may indeed make a person more likely to claim welfare but it might also by chance be associated with biological characteristics that make reproduction easier (for example, the employment-resistant personality profile may be associated with more plentiful eggs in women or more vigorously swimming sperm in men). Alternatively, it may be the case that it is poverty in general rather than welfare claiming in particular that is the key factor in determining number of children, as it is widely accepted that since about 1880 in developed Western nations, poorer people tended to have more children than wealthy people, even when there was no welfare state (the so-called demographic transition; Clark, 2007).

  Table 4.1 Average number of children under the age of 16 in working, mixed and workless households in England and Wales during April–June 2013

  Source: Labour force survey household dataset.

  Fortunately, these problems have been addressed by studies that explored effects of changes in welfare generosity on offspring numbers in welfare claimants and which show that increased welfare generosity increases the number of children born to claimants. This is the opposite pattern to that which we would expect if poverty increased reproduction. As an example, in the USA in 1996, the Personal Responsibility and Work Opportunity Reconciliation Act was introduced which ended the requirement for states to provide extra benefits to families who had additional children whilst on welfare, in an attempt to discourage families on welfare from having extra children (the so-called family cap). However, even before this act was signed into law, a number of states had received waivers from federal government that allowed them to cease providing extra benefits for extra children born to welfare claimants.

  This situation created a natural experiment which was utilised by Argys et al. (2000) to compare the effects on reproduction of states with differing policies on the payment of financial benefits for children born to 1,168 unmarried women who reported receiving income from welfare during at least one year between 1979 and 1991. Argys et al. (2000) were also able to contrast differing levels of generosity between states that did pay extra benefits for each additional child born to a mother on welfare. For example, Argys et al. (2000) reported that in 1991, a mother in Mississippi would receive $101 on the birth of a second child whereas a mother in California would receive $178.

  What Argys et al. (2000) found is that economic variables influence the reproductive behaviour of welfare-claiming women. The availability of incremental welfare benefits awarded for each additional child produced was positively and significantly associated with reproduction. This finding is congruent with those of other studies in the USA that show a similar tendency for increases in the generosity of welfare payments to be associated with increases in births amongst recipients (for example, Moffitt, 1998). However, Argys et al. (2000) additionally found that reductions in child-bearing in response to reduced welfare generosity were achieved by a decrease in pregnancies rather than an increase in abortions, concluding that: ‘capping benefits will lead to a reduction in births by making women more efficient contraceptors as opposed to increasing abortions’ (Argys et al., 2000, p. 584).

  Overall these results suggest a willingness on the part of the women studied by Argys et al. (2000) to manipulate reproduction via contraception in order to maximise returns from per-child welfare benefits. The credibility of their findings is boosted by their congruence with earlier work showing that reproduction in general is sensitive to tax subsidies and universal child benefits (Whittington, Alm & Peters, 1990; Whittington, 1992; Milligan, 2005) and the general finding that spells on welfare support tend to be longer in response to increases in welfare generosity (O’Neill et al., 1987).

  Evidence that welfare claimants tailor their reproductive behaviour according to whether it is financially worthwhile in terms of benefits claimed has also been found in the UK. In a bizarre coincidence, whilst the US government in the mid- to late 1990s was explicitly seeking to limit reproduction in welfare claimants by introducing the family cap, in 1999 the UK government boosted the generosity of per-child welfare payments by approximately 50 per cent in real terms. So generous were these welfare payments that the birth of a first child in a household in the bottom fifth of the UK income distribution would bring a cash benefit increase equivalent to a 10 per cent rise in net household income.

  The effects of this change to UK welfare provisioning have been studied in detail by Brewer et al. (2011), revealing that reproduction is more sensitive to changes in welfare legislation than had ever previously been shown: not only did these increases in the generosity of per-child welfare in the UK in 1999 increase the number of children born to benefit recipients by approximately 15 per cent, but also this effect was nuanced according to the specific opportunity-cost circumstances of the individual women. Brewer et al. (2011) were able to show this because the increases in benefits were accomplished by increases in income support (for workless households) and the introduction of the Working Families’ Tax Credit (WFTC), which resembles the US Earned Income Tax Credit (EITC) scheme, which was intended to increase financial incentives for welfare claimants with children to enter the labour force. Their analyses showed a bigger increase in births to women in couples than in single mothers, which reflected the effect of the WFTC rules that meant that mothers in couples whose partner worked 16 hours or more per week could leave the labour force altogether whilst nevertheless increasing their unearned income. Moreover, Brewer et al. (2011) found that the increase in reproduction in response was causally linked to the availability of more generous benefits, because they interviewed claimants and found that increases in reproduction were deliberately accomplished by discontinuation of contraception. This finding is in line with the earlier conclusion of Argys et al. (2000) that reductions in births in response to reductions in welfare benefits are accomplished by increased use of contraception.

  These findings have been backed up by evidence from a different source, namely two recent analyses by the Office for National Statistics in the UK. First, a statistical bulletin released in July 2015 shows that there were 695,233 live births in England and Wales in 2014, a decrease of 0.5% from 698,512 in 2013. The bulletin proposed that the reduced fertility levels witnessed in 2014 could have been caused, in part, by the reductions in the generosity of welfare benefits that were implemented by the UK government in 2013. This analysis is congruent with the notion that welfare claimants voluntarily limit their fertility in response to reduced generosity of benefits. Second, it was found that women from 27 European nations who were living in England and Wales in 2011 on average had 0.34 more children than their compatriots at home. Furthermore, these data showed stratification according to the prosperity of the country of origin that is congruent with the idea that women from poorer nations who move to live in England and Wales are adjusting their birth rate upwards to take advantage of the more generous child benefit payments in the UK compared to their home nations. For example, women from nations with approximately similar levels of affluence to the UK who were living in England and Wales tended to have a similar number of children to women in their home country. For example, women from the Netherlands living in England and Wales in 2011 averaged 1.71 children and those living in the Netherlands averaged 1.76 children. In contrast, women living in England and Wales who came from much less affluent European nations typically had significantly more children than women who remained in their home country. For example, women from Romania living in England and Wales in 2011 averaged 2.93 children whereas those living in Romania averaged 1.25 children (source: Childbearing of UK and non-UK born women living in the UK – 2011 Census data; Office for National Statistics, 2014).

  Viewed as a whole, these studies on reproduction and welfare seem to present two paradoxes. First, raising children generates work in its own right, so it might seem plausible that, instea
d of having extra children in order to increase welfare benefits, employment-resistant claimants should be deterred from doing so by the prospect of 18 years or so of childcare. Second, increases in household income due to greater welfare benefits should reduce disadvantage and therefore reduce risk of personality mis-development amongst the children of claimants.

  Both these paradoxes can be resolved by remembering that in Chapter 3 we saw that children with relatively low self-control also turn out to have an increased risk of becoming teenage parents, suggesting that part of the employment-resistant personality is a lack of foresight. This suggests that employment-resistant claimants would indeed be more likely than average citizens to have extra children in response to welfare incentives, because their lack of foresight means they will focus on the short-term reward of increased child benefits and pay relatively little attention to the longer term reality that raising each extra child means 18 years of extra work.

  The link shown in the Dunedin Study between the employment-resistant personality profile and rapid, irresponsible reproduction is confirmed by evidence from the study of problem families by Tonge et al. (1975). This showed that the problem families not only had significantly worse work records than the comparison families, but also had significantly more children. There were 200 offspring in the 33 problem families compared to 132 in the comparison families. More than 50 per cent of the problem families (17 families) had six or more children. In contrast, only nine comparison families had six or more children.

  In the final chapter of their book, Tonge et al. (1975) listed the following four values as underpinning the maladjusted behaviour (including reproductive patterns) of the problem families: 1) rules were ignored; 2) discomfort was ignored; 3) long-term consequences were ignored; and 4) education was distrusted. The notion that the problem families were chasing extra welfare funds by having extra children, regardless of the long-term consequences, was echoed in the qualitative impressions of the researchers: for example, the researchers found that most of the problem families typically received sufficient welfare payments to cover the cost of such key necessities of life as food, clothing and heating. Yet instead of carefully husbanding their welfare money to ensure that their children were adequately provided for, the problem families tended to waste it on such needless items as expensive toys, luxury chocolates, cigarettes and alcohol, a spending pattern that typically left their children poorly fed, poorly clothed and in unheated houses.

  This impression was supported empirically by the findings of Tonge and colleagues who compared the protectiveness of the mothers: whereas in the comparison families, 30 mothers were rated as normally or over-protective of their children, only ten of the problem family mothers gained this rating. In contrast, 20 mothers in the problem families were rated as under-protective or neglectful of their children, compared to just one mother in the comparison group. These data on protectiveness of mothers towards their children were backed up by the child mortality rates (per 1000 live births), which were 66 in the children of problem families and 39 in those of comparison families.

  If we return for a moment to the ecological principle of r–K selection, this finding supports the notion that the employment-resistant personality profile biases the individual towards r selection (rapid, irresponsible, high-volume reproduction with little effort put into raising the offspring). Furthermore, by removing the need to compete for resources, the welfare state hands the reproductive upper hand to individuals who are biased towards r selection: individuals who, I argue, possess the employment-resistant personality profile. Additionally, the research of Tonge and colleagues suggests that simply throwing more money at the problem families would increase the frequency of unnecessary, impulse purchases (for example, electronic gadgets and luxury chocolates) but that the extra funds would not necessarily improve the situation of their children in any meaningful economic or social sense.

  This pattern of irresponsible spending that is displayed by problem families is perhaps summed up best in a passage from Chapter 2 of their book, which addressed the challenge of finding the families:

  At first glance it is difficult to single out problem families from others on the estates. All the houses are structurally similar, either terraced or semi-detached. They appear as endless rows of grimy brick and pebbledash, with doors painted in the colour chosen by the Public Works Department of the Corporation. On a grey day all merges imperceptibly with the horizon. On closer inspection the levelling effect of Corporation maintenance does not disguise everything. Here and there a front door knob is surrounded by black greasy marks, a window pane is broken or boarded up and curtains are half closed. Somehow the paint has weathered less well than on neighbouring houses and the front garden is characteristic. An overgrowth of natural flora competes with broken bottles, sodden cardboard and the rusting remains of once-expensive toys, prams, cycle parts and other scrap. A well-worn earth path leads from the dominant door to the pavement by the most direct route which involves a hole in the hedge or a section of shattered fencing.

  (Tonge et al., 1975, p. 11)

  Conclusion

  Organisms can optimise their fitness either by having many offspring, but investing little effort in their raising (r selection) or having fewer offspring and conscientiously nurturing them (K selection). The utility of these strategies is influenced by resource availability: when resources are freely available, r selection is optimal, but when resources must be competed for, K selection is optimal. In humans, individuals with personality characteristics approximating to the employment-resistant personality profile are biased towards r selection. In pre-welfare times (at least when the geo-political situation was stable, such as in pre-Industrial Revolution Britain), employment-resistant individuals appear to have been at a reproductive disadvantage owing to their problem of maintaining an income, but since then, the employment-resistant personality profile has gained the reproductive upper hand. I argue that one contributory factor to this shift in fitness from K selection to r selection in modern humans is the reduction in competition for resources that is brought about by the advent of the welfare state.

  5

  Childhood Disadvantage and Employment-Resistance

  The next three chapters of this book summarise evidence on the transmission of personality characteristics from parents to children. This theme is important because, if it was the case that personality characteristics were not transmitted between generations, then a welfare state that boosts the number of children born to employment-resistant claimants would not affect the personality profile of the population. The child of a welfare claimant would be just as likely to become a conscientious and agreeable ‘solid citizen’ as the child of an employed person.

  In the previous chapter, we saw evidence that individuals with the employment-resistant personality profile are especially likely to maltreat their children. In this chapter, we shall see evidence that child maltreatment is a significant cause of employment-resistant personality characteristics. Viewed as a whole, these two findings point to the existence of a trans-generational cycle of maltreatment, in which today’s maltreated children become tomorrow’s employment-resistant perpetrators of child maltreatment. More generally, it seems that employment-resistant personality characteristics can be transmitted from parents to children by a purely environmental mechanism, namely exposure to maltreatment as a child.

  When considering the topic of child maltreatment, it is natural to focus on extreme, headline-grabbing cases, such as when a child spends years chained up alone in a darkened basement, or is used as a sex slave, or injected with heroin to quieten him or her down. These children would be expected to suffer personality mis-development, yet such cases are thankfully extremely rare and cannot plausibly account for personality mis-development on a large enough scale to cause significant alterations to human capital.

  In contrast, milder forms of maltreatment, such as parental neglect or inattention, are unlikely to make tabloid headlines and s
o may be overlooked as a cause of personality mis-development. Yet there is evidence that this milder type of maltreatment can impair child development. Moreover, as we saw in the previous chapter, research by Tonge and colleagues suggests that child maltreatment of the passive type – what I shall henceforth refer to as child neglect – is significantly more common amongst welfare-claiming, employment-resistant parents than it is amongst employed parents, even though the latter have less free time than the former. The first aim of this chapter is to summarise studies supporting these claims.

  However, suggestions are not enough to prove a scientific point: the studies I have cited so far do not tell us conclusively that neglect has a significant role in causing children to turn out to be more employment-resistant than children who have not been neglected. Even if they could do this, these studies still cannot disentangle whether it really is child neglect that is driving this effect, as opposed to genetic inheritance. Nor can such studies eliminate the possibility that it is poverty that is driving this effect, rather than being a welfare claimant per se. To answer these questions we need to turn to research that combines a longitudinal design with proper experimental controls. In other words, we need to look at experiments in which disadvantaged children are randomly assigned to receive in loco parentis intensive preschool tutoring and then their life outcomes as adults are systematically compared to those of children who have not been given such tutoring.