The Welfare Trait
Page 18
The true key to non-circular corroboration of the welfare trait theory is the insight that an employment-resistant attitude is just one manifestation of a general tendency to behave in an antisocial, norm-breaking manner. As we saw in Chapter 4, another manifestation of this personality profile is a willingness to have extra children in order to increase welfare income, but then to neglect those children. According to my theory, it is this latter facet of the employment-resistant personality profile that gives the welfare state its special power to alter the personality make-up of the nation by boosting the number of children who are born into disadvantaged households and who suffer personality mis-development as a result.
If it is true that the welfare state is causing the employment-resistant personality profile to proliferate, we would therefore expect workless families not only to display high rates of antisocial behaviour, but also to have more children than average – children who are then neglected. Evidence of this type has already been provided in earlier chapters by the longitudinal studies of problem families undertaken in Sheffield by W. L. Tonge and colleagues, but the sample sizes in those studies were relatively small (66 families in total) and they were completed more than 30 years ago.
What we require is a larger, more recently collected data set that can be used to provide independent replication or refutation of the findings of the Sheffield studies. Just such a data set has been collected in the UK by the Troubled Families Programme, which was initiated in December 2010 with the aim of helping families with three or more adjustment problems to turn their lives around. As we shall see, these data replicate the results of the Sheffield studies, as they show that troubled families not only display higher than average rates of antisocial behaviour, but also have more children than average families. Moreover, those children tend to be neglected. The former association is unsurprising, since unemployment and antisocial behaviour were two of the criteria used to select families for the Troubled Families Programme. However, family size was not one of the entry criteria for the programme. This means that the data from the Troubled Families Programme provide independent support for my hypothesis that the welfare state is proliferating antisocial personality characteristics by boosting the number of children born into neglect-prone households.
Continuing the theme that employment-resistance is just one manifestation of a general tendency to behave in an antisocial manner, a particularly well-documented form of antisocial behaviour is crime. This opens up a third avenue for corroboration of the welfare trait theory: if it is true that the welfare state is modifying the personality profile of the population towards greater employment-resistance, then we should not only see a reduction in work motivation from generation to generation following the introduction of the welfare state, but also find that there is a concomitant upsurge in crime as each generation passes. More specifically, because antisocial personality characteristics have a particularly strong link to criminal violence (Hodgins, 2007; Jüriloo et al., 2013), we should see an upsurge in violent crime, beginning about 15 years or so after the implementation of the welfare state, as the first generation of individuals with mis-developed personalities grow old enough to begin showing up in the crime statistics. In the final part of this chapter, I summarise evidence that supports this hypothesis.
Effects of the welfare state on work motivation
When assessing the effects of the welfare state upon work motivation, we are fortunate to be able to call upon decades of research on welfare-usage trends in Scandinavian countries. These nations have become the leaders of such research because they typically implemented a modern welfare state before most other nations of the Western world and economists therefore have been able to study its effects for longer. At the forefront of attempts to understand the effects of the welfare state on work motivation is the work of Assar Lindbeck, who, as mentioned in Chapter 1, proposed that the welfare state risks destroying itself via the erosion of norms connected to work and responsibility. In Lindbeck’s view: ‘The basic dilemma of the welfare state, however, is that the more generous the benefits, the greater will be not only the tax distortions but also, because of moral hazard and benefit cheating, the number of beneficiaries’ (Lindbeck, 1995, p. 2).
Central to Lindbeck’s work is the notion that generous welfare benefits have a delayed effect on behaviour owing to the time taken to erode the social norms of previous generations. There is some empirical support for this idea, since it has been found that generous welfare benefits weaken parents’ incentives to instil work-motivation norms in their children (Lindbeck & Nyberg, 2006). A delayed negative impact of the welfare state on the functioning of society is also consistent with the welfare trait theory, which can explain it in terms of the time taken for claimants’ children who have suffered personality mis-development due to childhood disadvantage to grow old enough to show up in government statistics on unemployment, crime and other key metrics of societal dysfunction. Either way, the danger of welfare-induced reductions in work motivation is clear: ‘if we do not watch out for hazardous dynamics, there is a risk that the welfare state will destroy its own economic foundations’ (Lindbeck, 1995, p. 2).
Basic evidence to corroborate the notion that the welfare state can erode work motivation has been provided by studies that compare unemployment rates and welfare generosity between countries. For example, Nickell (1997) investigated the effects of welfare benefits on unemployment rates in 20 OECD countries, namely Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the UK, Canada, the USA, Japan, Australia and New Zealand. This study showed that high unemployment was significantly associated with generous welfare benefits, as well as with lengthy/unlimited periods of entitlement.
But studies that compare unemployment rates and welfare generosity are something of a blunt instrument, since they do not directly measure work motivation and there are many other reasons why unemployment rates and welfare generosity might be connected. A more specific empirical test of Lindbeck’s theory is to use self-report data on attitudes to welfare usage. For example, Heinemann (2008) used data from the European and World Values Survey to investigate associations between benefit morale and welfare generosity in 33 OECD countries. Benefit morale is a construct that reflects willingness to claim benefits to which the individual is not entitled: the more willing the individual is to cheat the welfare state, the lower their benefit morale. In four time periods (1981–1984, 1989–1993, 1994–1999, 1999–2004), there was a general increase – with a substantial time lag – in the willingness of the citizens of these nations to make fraudulent benefit claims. This trend tended to occur more strongly in more generous welfare states, such as Sweden. In order to control for age effects, Heinemann (2008) also examined differences in benefit morale between individuals from different birth cohorts who were surveyed at the same age, finding that younger birth cohorts were more willing to cheat the welfare state than older birth cohorts, even though their age at the time of survey was the same. This latter finding is especially supportive of the welfare trait theory because if the welfare state is shifting the personality profile of the population towards lower levels of conscientiousness and agreeableness, we would expect to see that young cohorts would be more comfortable than older cohorts when fraudulently claiming welfare handouts.
Subsequent studies have mostly produced similar results to those obtained by Heinemann (2008). For example, Halla, Lackner and Schneider (2010) also examined data from the European and World Values Survey, but used different statistical analyses. They studied effects of welfare generosity on benefit morale over three different time lags: short run (the effect of current generosity on current benefit morale), medium run (the effect on current benefit morale of increased welfare generosity five years previously) and long run (the effect on current benefit morale of increased welfare generosity ten years previously). They found that high welfare spending in the current period had a small pos
itive effect on current benefit morale. In the medium run (five years) there was no effect, but in the long run (ten years) there was a significant negative effect, with a percentage point increase in welfare generosity followed ten years later by a decrease in benefit morale of approximately 0.22 points.
This finding backed up Heinemann (2008), with Halla and colleagues explaining it by suggesting that ‘individuals have to experience generous welfare arrangements for quite some time until they adapt their social norm towards accepting benefit fraud, or at least considering it to be a minor offence’ (Halla et al., 2010, p. 69). This explanation sees individuals as responsive to the perverse incentives of the welfare state, but with a delay in their willingness to defraud the welfare state that is caused by their need to overcome their ingrained reluctance to cheat the system. However, this result is also consistent with my theory of welfare-induced personality mis-development which explains the delayed action of benefit increases by postulating that faulty norms are acquired during childhood (for example, by observing a parent committing benefit fraud) that then make the individual more likely to cheat the welfare state a decade or so later when they themselves become old enough to start claiming benefits.
Halla and colleagues also found that, after statistical refinements which were not performed by Heinemann (2008), there were no significant differences between birth cohorts in willingness to cheat the welfare state. This finding counters the welfare trait theory, which predicts a reduction in work ethic from generation to generation due to a progressively greater frequency of individuals with the employment-resistant personality profile in each birth cohort. However, interestingly, Halla and colleagues also showed that labour-market status modulates benefit morale: unemployed individuals had a significantly lower level of benefit morale than employed individuals (approximately four per cent lower). Lower morale could be a product of unemployment, but it could also be a cause: this latter possibility is consistent with the welfare trait theory, on the basis that employment-resistant individuals are less motivated than average citizens to work for a living and so tend to be unsatisfactory employees, causing them to filter into the ranks of the unemployed at higher rates.
A recent study that could be portrayed as countering the welfare trait theory investigated the relationship between employment commitment and welfare generosity in 18 European countries (van der Wel & Halvorsen, 2015). This study used data from the European Social Survey which included the question: ‘I would enjoy having a paid job even if I did not need the money.’ Participants responded by endorsing one of the following options: ‘Strongly agree’, ‘Agree’, ‘Neither agree nor disagree’, ‘Disagree’, ‘Disagree strongly’ and ‘Don’t know’. The response options ‘Strongly agree’ or ‘Agree’ were coded as one and the remaining options were coded as zero, except for ‘Don’t know’ responses, which were coded as missing.
Using responses to this question as a measure of work motivation, van der Wel and Halvorsen (2015) found that average levels of commitment to employment were high in more generous welfare states. For example, the highest levels of employment commitment were found in Norway, Denmark, the Netherlands and Switzerland, which were also the most generous nations in terms of welfare benefits. However, unlike the previously mentioned research, this was a cross-sectional study and therefore is unable to shed light on causation. For instance, the moral hazard of generous welfare benefits may be lower in countries with a high baseline level of work motivation than in countries with a weak work ethic. Furthermore, the cross-sectional study design cannot detect change in work motivation from generation to generation, which is a key part of Lindbeck’s warning concerning the self-destructive nature of the welfare state.
However, the most serious criticism of this study (and much of the research in this field) is that a single, self-report question was used to measure work motivation. It would be prudent therefore to test the validity of these findings by examining whether welfare-claiming behaviour showed similar patterns to self-reported attitudes. The work of Martin Ljunge (2011) provides just such a test. First, instead of looking at changes in attitudes to welfare usage, Ljunge (2011) used data on longitudinal changes in actual welfare usage in Sweden. Second, Ljunge (2011) selected sick-benefit claims as the focus of his research because, for the first week of each spell of sick leave, the Swedish welfare state allows the individual to decide for themselves whether they are sick. Ljunge (2011) describes this system as being ‘like a giant marshmallow test’, since its laissez-faire design gives unconscientious individuals a free hand to defraud the welfare state by pretending they are sick when they are not. This impression is supported by previous research which showed that the number of men who reported sick in Sweden increased significantly when there was a major televised sporting event (Skogman Thoursie, 2004). The uptake of sick benefits in Sweden therefore allows objective measurement of changes in work motivation between generations.
Ljunge (2011) used data for Swedes aged between 22 and 60, randomly sampled from the 1974 population and followed for 17 years. Ljunge (2011) selected this time period because the rules governing sick leave remained constant between 1963 and 1990, with data on sick leave becoming available from 1974 onwards. This means that any changes in the usage of sick benefits over this time cannot be attributed to changes in welfare rules.
In line with the welfare trait theory, as well as the findings of Heinemann (2008), Ljunge (2011) found that younger generations of Swedes claimed more sick-leave benefits than older generations. The effect was substantial, with each generation having approximately a one per cent higher take-up of sick benefits than the previous generation. For example, the generation born in 1920 on average claimed sick benefit in fewer than half of the years that they were in the labour force (45 per cent) whereas the generation born in 1960 had a take-up rate of almost 80 per cent, making a sick-benefit claim in eight out of every ten years of their working life.
This difference between generations in sick-benefit claims cannot be explained by health differences, since younger generations of Swedes had better health than older generations and so, if they were being honest, they should have claimed less sick benefit than their older compatriots. Moreover, Ljunge (2011) found that these inter-generational differences in sick-benefit claims were mirrored by differences in unemployment-benefit claims: younger generations claimed more unemployment insurance than older generations.
Ljunge (2011) then tested whether these changes in welfare usage were mirrored by attitude changes. He used data from the integrated European Values Survey and World Values Survey (EVS/WVS), focusing on data from 95 countries in five waves of collection from 1981–1984 to 2005–2006. In line with Heinemann (2008) and his own data on sick-benefit uptake, Ljunge (2011) found that younger generations of Swedes viewed it as more acceptable to claim benefits than older generations of Swedes. Furthermore, he found that this change was echoed in most of the other 95 countries studied.
Ljunge (2011) suggests that the trend for younger generations to abuse the welfare state more than older generations could be explained by increased exposure to role models who have a high take-up rate of welfare benefits. Ljunge’s analysis is consistent with the welfare trait theory, but leaves unanswered the question as to what causes this increased exposure to dysfunctional role models. The welfare trait theory can explain this, on the basis that a welfare state which boosts the number of children born into disadvantaged households will increase the number of children exposed to dysfunctional, employment-resistant parental role models.
Viewed as a whole, the balance of evidence provided by these economic studies fits the notion that the welfare state is increasing the frequency of individuals with mis-developed, employment-resistant personality profiles – profiles that make them willing to defraud the welfare state. However, since the welfare trait theory was partly inspired by these same Scandinavian economic studies, any attempt to use them to corroborate it is an exercise in circularity. To begin the proc
ess of breaking this circularity, we now move on to independent evidence concerning the capacity of environmental factors to mould personality.
Cross-cultural comparison of social attitudes
Differences in social attitudes can be measured using standardised tasks that model important everyday interactions. One such task is the Ultimatum Game (UG), which measures attitudes that correspond approximately to the personality domain of agreeableness, such as fairness, generosity, altruism, unselfishness and cooperativeness. It involves two players whose identities are concealed from each other. The anonymity of the players is a key feature of the UG because it prevents their behaviour being influenced by pre-existing relationships, such as friendship or genetic kinship.
One UG player (known as the proposer) is allotted a substantial asset that can easily be divided (for example, a sum of money equivalent to two day’s wages in the society in question) and is instructed to offer a part of the asset to the second player (known as the responder). The responder, who is aware of the size of offer, as well as the overall size of the asset, is free either to accept or reject the proposer’s offer. If the responder accepts the offer, he or she receives that part of the asset and the proposer receives the remaining amount. If the responder rejects the offer, then neither player receives any of the asset.
The UG is a one-shot game: once the response has been made, the game terminates and the two players depart with their allotted winnings in the event of an accepted offer, or with nothing if the offer is declined. Importantly, the UG requires no work or material contribution from either player – the asset is, in effect, a free gift. This latter feature makes the UG relevant to discussions of a welfare state in which benefits are provided without requiring work in return.