The search for an ‘asset-effect’: What do we want from asset-based welfare?

Introduction

In an era of deep welfare retrenchment, it is perhaps unsurprising that recent steps towards the greater distribution of wealth and property have, apparently, been stopped in their tracks. One of the very first acts of the coalition government in Britain was to disband the internationally admired Child Trust Fund, and to halt the roll-out of the Saving Gateway scheme. Many viewed these policies as a first step towards a ‘progressive’ asset-based welfare regime; one in which there was a more egalitarian and fairer distribution of the wealth of an enormously affluent society, as well as offering an opportunity for many poorer households to build a financial cushion against the insecurities of the modern economy. These policies were therefore part of a movement that, at times, invited us to think about the welfare state differently: as a means of spreading wealth and opportunity, rather than just a safety net or system of mutual insurance (Sherraden, 1991). The loss of these policy initiatives is therefore a loss not just for their beneficiaries, but also for a more sophisticated model of welfare provision in a modern economy.

Yet it was never entirely clear what the wider movement for asset-based welfare (ABW) really hoped to achieve, so the wider ‘loss’ of ABW policies is inevitably hard to come to terms with. At times ABW has appeared to be a commonsensical but relatively limited means of smoothing welfare consumption for poorer households, whilst at other times it serves the more ambitious ideal of social inclusion and greater equality of social status (Prabhakar, 2008). Typically, this more ambitious ideal comes with the suggestion of an ‘asset-effect’ – the notion that assets have a positive effect ‘over and above’ their instrumental value. As we shall see, this is the subject of considerable empirical controversy. But it also leaves us with the crucial normative questions: what ‘effect’ do we want to see, in what ways do we want asset-ownership to transform the outlook and behaviours of individuals? It is the purpose of this paper to offer a clearer and more explicit analysis of these normative questions, as a first step in understanding the deeper purpose and policy implications of competing accounts of asset-based welfare and, crucially, the way in which we think about the rights and duties of citizenship in a modern welfare state.

The order of argument

The following sections set the policy and political context of our discussion, paying particular attention to the ABW aspirations of the current coalition government in Britain. The aim here is to highlight the ways in which the language of ‘asset-based welfare’ can easily be co-opted by a policy and political agenda based on the ideals of self-sufficiency and a minimal welfare state, rather than on ideals of social inclusion and strong welfare rights. Section 3 moves on to examine the instrumental case for ABW. The argument here is that whilst there is a strong case to be made for a relatively modest ABW agenda that helps poorer households build up savings and assets as a means of cushioning income shocks, we need to take greater care in assuming that much larger assets (such as homeownership) have the same positive impact on household finances and well-being.

In sections 4 to 6 the paper turns to the evidence for a deeper ‘asset-effect’, in which asset-ownership is more than an instrumental good. The argument in this section is that current claims for an asset-effect can in fact be reduced to more modest instrumental explanations of the value of ABW, and that, furthermore, we need significantly greater conceptual clarity regarding the nature of any ‘effect’. In particular, as I shall argue in section 7, this requires a sharper account of the relationship between ownership and psychological or behavioural transformation. Finally, sections 8 and 9 return to the question of ABW and citizenship, and argue for an account of social inclusion that recognises the normative importance of asset-ownership without having to rely on an ‘effect’, and without playing into the hands of a competing account of citizenship as a duty of self-sufficiency.

A hidden ‘asset-based welfare’ state

During the 2000s there was a concerted drive to create an ABW regime for poorer households in Britain. This agenda was, in large part, driven by ideals of social inclusion and fairness, as well as a practical means of tackling income poverty. We all know that this has come to a premature end, with both the Child Trust Fund (a universal endowment for all British children) and the Saving Gateway (a matched saving scheme for lower-income households) falling prey to the very first austerity cuts.

Yet there is still a very strong ‘hidden’ asset-based welfare regime; a system of rewards and subsidies that are highly favourable to Britain’s middle classes – and which sit uneasily with the ‘progressive’ ABW policy discourse, driven in large part by the ideal of social inclusion that motivated the Child Trust Fund and the Saving Gateway.

In Britain, for example, contributions to private (occupational or personal) pension schemes are tax free, and (in the case of defined contribution workplace schemes) neither employee nor employer has to pay National Insurance Contributions on pensions contributions. These exemptions, according to recent analysis in the Mirrlees Review, amount to more than a tax exemption; they are a positive subsidy, with an effective tax rate on employee contributions of minus 8 per cent (Mirrlees et al., 2011; Hills and Glennerster, 2013: 177). Moreover, pension tax relief is set by the income tax band of the saver and thus highly regressive: those in the 20 per cent lower tax band therefore receive far less tax relief than higher earners in the 40 per cent tax band.

Similarly, the government incentives for tax-free saving in Individual Savings Accounts (ISAs) reach a relatively wide proportion of the population in terms of income scale, but inevitably favour those with more to invest, meaning that the wealthiest in effect get the largest tax break. In the year 2010–11 there were approximately 11 million ISA accounts held by individuals with an annual income of between £10k and £30k, but the amount saved is relatively small (roughly £15k) compared to the much higher savings of the smaller groups with incomes above £30k. This pattern continues up the income scale: those earning above £50k having ISA savings of £25k, and even up to £45k for those earning £150k a year (HMRC, 2013).

There is also a vast subsidy for homeownership – the lodestone of narratives of ‘property-owning democracy’ – in Britain. In Britain owner-occupiers pay no capital gains tax on the increasing value of the asset, no longer pay tax on the imputed rental value of their home, and pay no tax on the land. There have also been direct subsidies to encourage greater homeownership, most notably the sale of council homes to tenants under the 1980 Right to Buy legislation. From 1980 to 2009 over 2.5 million properties were sold at discounts to the open market rate of up to 60 per cent for houses and 70 per cent for flats. Recent analysis suggests that the accumulated value of all these discounts is somewhere between £150 to £200 billion in total (Hills and Glennerster, 2013: 187). Whilst the Labour government capped the cash value of subsidy in 2005 at a maximum of £38,000, the current government has increased the cap to £75,000 across England. Even more recently the coalition government has pushed ahead with a Help to Buy scheme, sometimes justified as a Keynesian stimulus, but in reality a new plank in a ‘property-owning democracy’ political narrative (Financial Times, 2013).

This is only the briefest sketch of the hidden ABW regime in Britain. A fuller account would include recent policy developments that are not directly framed in terms of ABW, but which nevertheless highlight the relationship between wealth and a range of important outcomes. Of particular salience is the recent identification of wealth inequality as an underlying driver of health inequalities (Marmot Review, 2009), and (more controversially) of social malaise, for example, higher levels of crime and drug abuse (Wilkinson and Pickett, 2010). At the same time, there are growing calls to use accumulated personal wealth, mostly housing wealth, as a partial solution to health inequalities later in life, especially as a means of funding social care (Dilnot, 2011).

But the broader purpose of this paper is to examine the normative foundations of ABW, especially its relationship with different and competing accounts of ‘citizenship’, rather than to enter into a detailed policy analysis. These wider policy issues (for example, the relationship between wealth, social care and health) will, in any case, raise the same normative issues that are discussed in this paper.

It is important, therefore, to realise that the rise of state sponsored ownership – and, to an extent, ‘hidden’ ABW more generally – is underpinned by strong normative assumptions as well as social policy goals; ownership is seen not just as a right or an opportunity, but also a duty – and, crucially, a means of teaching individuals to be better citizens. It is a duty because, at first sight at least, it requires the individual to ask less of the state and her fellow-citizens. Ownership is didactic too: it changes our outlook and attitudes. Thus, according to Margaret Thatcher, council houses were ‘breeding grounds of socialism, dependency, vandalism and crime’; whereas homeownership taught ‘all the virtues of good citizenship’ (Campbell, 2003: 234). And in speaking of the wider programme of privatisation, the deeper purpose of the very particular brand of ABW was made explicit; the aim was to use the state as an educative tool to ‘change the soul’ (Sunday Times, 1981). Recent policy narratives reflect the same assumptions. Iain Duncan Smith also thinks that homeownership can create good citizens:

[W]e want to encourage tenants’ aspirations: we propose rewarding social tenants’ constructive behaviour with equity shares in their homes. With the privilege comes the responsibility of maintaining one’s own assets; and in the long run this will transform our estates. (Duncan Smith, 2008: 6)

And a similar prescription is endorsed by David Cameron:

Generations of families are trapped in social housing, denied the chance to break out or to buy their own property. I don’t want a child’s life-story to be written before they’re even born, and a responsible housing policy which helps people up and out of dependency can help re-write that story. (Cameron, 2009: 3)

What do we want from asset-based welfare?

This is not a story that can be ignored by ‘progressive’ or social inclusion advocates of ABW. The story cannot be ignored because progressive accounts share the same crucial assumption that asset-ownership can have a transformative effect on the individual, bringing about not just behavioural changes, but also changes in the individual’s sense of self and their place in society. What those changes amount to, however, is deeply contestable, and without greater clarity an ostensibly progressive account of ABW runs the serious risk of legitimising policy and practices that are better suited to a small-state Conservatism (or minimalist liberalism).

There are two key categories of contestation. Firstly, there is the empirical question of what kind of behavioural changes are in fact brought about by ownership. Secondly, there is the normative question of the degree to which such changes are desirable. This second question draws us on to the ground of basic principle, forcing an articulation of (for example) where advocates of ABW place themselves on questions of equality and citizenship. Where one account may be strongly egalitarian and see asset-distribution as a right of ‘citizenship’ based on something like equality of social status, another account may see asset-distribution as a means of ‘rewarding’ certain types of behaviour that distinguish ‘good’ from ‘bad’ citizens.

These tensions have, I suggest, received far too little attention in the literature, and I hope to take a small step in redressing this towards the end of this article. But the immediate point is that the idea of an ‘asset-effect’ is not one that can be exclusively claimed by any particular ideological orientation or political position. Progressives – those who want to use asset-ownership as a vehicle of social inclusion rather than social hierarchy and exclusion – need to recognise this if they do not want the whole ABW agenda to be co-opted for inegalitarian purposes.

There is a further interesting anomaly in the current ABW debate. The association between citizenship and asset-ownership is particularly strong in the case of housing policy, yet it has received very little attention in the literature on ABW, which in the UK has been dominated by analysis of relatively small amounts of wealth. This is perhaps understandable when we consider the social policy objectives of the ABW agenda, which has been aimed at poorer households who may never own their home. Nevertheless, it is a risky strategy in social policy terms. It is risky, at the most basic level, because it neglects the extent of poverty and disadvantage in owner-occupied households: roughly one quarter of UK households living in poverty are of working age and living in owner-occupied households with a mortgage (Gregory, 2011). But the lack of engagement with the issues of homeownership is also risky in normative terms, because – as we shall see – it encourages us to neglect the variant of ‘citizenship’ that typically accompanies the type of ‘hidden’ ABW I have described; a vision of citizenship that is unlikely to fit the normative intuitions of many advocates of ABW as a ‘progressive’ social policy.

Owning-up: The evidence base

First, however, for the (limited) evidence base on the impact of ABW policies. Thus far the evidence is seemingly quite positive. In the United States, an early evaluation of the Individual Development Account scheme (designed to help low-income households save) found that IDA participants self-report more confidence about their futures (93%), felt more economically secure (84%) and felt more in control of their lives (85%) because they have IDAs (Moore et al., 2001). Positive outcomes have also been reported in terms of greater participation in homeownership and business ownership.

Yet more recent reviews of the evidence suggest that these and subsequent positive findings are methodologically problematic: impact studies are based on volunteer participation and thus risk a selection bias (with the additional problem of potentially inaccurate self-reporting), with little actual provable improvement in financial well-being. Moreover, the evidence to support the assumption that IDAs play a causal role in increasing homeownership and college enrolment is ‘virtually nonexistent’ (Richards and Thyer, 2011: 360). Another review of the evidence also points to the absence of longitudinal evaluations of IDAs and, tellingly, very limited and ‘primarily descriptive’ attempts to measure the psychological impact of participation in IDAs (Zielewski et al., 2009: 13).

In the UK there has been a great deal of reliance on two longitudinal studies. The first of these studies (Bynner, 2000; Bynner and Despotidou, 2000; Bynner and Paxton, 2001) is intergenerational and draws on data from the National Child Development Study, which has followed a random sample of children born in one week in 1958. Controlling for factors such as parental education and occupation, this study found positive associations between small asset-holdings (as little as £300 in 2003 values) at the age of 23 and improved outcomes in health and employment. A slightly later study used the same methodology to successfully replicate the results using data from the British Household Panel Survey (BHPS) (McKay and Kempson, 2003), but could not replicate the results with a different methodology, once other controls (for example, ethnicity or early marriage) were added. Significantly, even without these extra controls, there ‘were no significant effects of assets … on most of the citizenship measures’ (McKay and Kempson, 2003: 64).

Yet, at the same time, this critical response has itself been criticised on similar grounds: BHPS data can only yield correlations, not causes, so (for example) it could be that non-smokers are more likely to acquire assets at 23 in the first place, rather than the assets are ‘preventing’ them from taking up smoking later in life (McKnight, 2011). More recent analysis, mindful of the temptation to promote correlations as causes, has cautiously found an association between low levels of asset-ownership (roughly £1000 in 1980 prices) early in an individual’s life and improved outcomes later on life; for example, higher employment rates, higher rates of pay within employment, and positive outcomes in terms of physical and mental health (McKnight and Karagiannaki, 2013). But whilst these more recent studies are rightly cautious about inferring causation from these findings, they do go on to discuss asset-effects as a real phenomenon and do in fact offer causal explanations; for example, the hypothesis that wealth allows access to services that contribute to psychological well-being (McKnight and Karagiannaki, 2013: 138).

The crucial conclusion for our purposes, however, is not a technical point of order about methodology; it is simply that the very idea of an asset-effect, as yet at least, has very little empirical weight behind it. It is also important to recognise that the evidence we have been offered is limited to small amounts of wealth (below £1000) and that it would be an unwarranted leap to assume that the same evidence is valid in the case of much larger assets, such as a home. And though this remains an under-researched phenomenon, there are strong intuitive reasons to think that the relationship between assets and risk might be reversed once we move beyond a certain value, once we are more anxious about what we stand to lose rather than gain, and particularly when the loss has the emotional attachments of a home (Gregory, 2011).

This leaves us with a range of important empirical questions to be asked of asset-ownership and risk – and a whole research agenda to be mined. In the context of this paper, the provisional conclusion we can draw at this stage is that the instrumental value of housing as an asset is insufficiently understood. Too often we simply assume, for instance, that homeownership is an unalloyed good.

But the broader point is that we need to not just garner robust and consistent evidence on the instrumental value of assets, but to do so in a way that respects the boundaries of different types of justification. In the following section I argue that the search for an ‘asset-effect’ requires a coherent distinction between the instrumental and ‘intrinsic’ (or ‘pure’) benefits of asset-ownership, and I turn to this in the section after that. This, as we are about to see, is of particular importance when we turn to the claim that asset-ownership can have an ‘effect’ over and above the goods they give access to.

The conceptual journey to an ‘asset-effect’

As I will go on to introduce a number of other concepts and distinctions, it is useful to summarise them here, in order of appearance, before the substance of the argument to come. First we need to draw a distinction between:

Within the category of ‘intrinsic effects’ a further distinction is then drawn, between:

We shall move through this structure as the argument proceeds. One final distinction needs mentioning, however. Rajiv Prabhakar has drawn a useful distinction between ‘social policy’ and ‘citizenship’ accounts of asset-based welfare (Prabhakar, 2008). Roughly speaking, ‘social policy’ maps onto my ‘instrumental’ category. ‘Citizenship’ accounts, conversely, are based on a right to assets and property. In principle, we could find accounts that are either one or the other. In practice, however, most accounts are likely to contain elements of the two. Prabhakar himself points to the inherent normativity of ‘social policy’ narratives, which typically come with an account of welfare dependency or incentives and behavioural changes. Indeed, if the argument of this paper is correct, a ‘citizenship’ component is inevitable in any account that can properly claim the title of ‘asset-based welfare’. The further distinctions introduced in this paper are thus designed to help to distil an important ambiguity in some recent research, which tends towards an equivocation between the instrumental value of ABW and a more normative (or ‘intrinsic’) justification.

Going above and beyond: Instrumental versus intrinsic value

Here I address the first fundamental distinction outlined in the ‘Conceptual journey’: the distinction between instrumental and intrinsic effects. This distinction is not always a clean one. We can see this by looking at an instrumental example. The Saving Gateway matched savings scheme had a positive outcome on the ability of many households to withstand income shocks, helping to prevent a spiral of debt and poverty when (for example) income has fallen but a child still needs a new school uniform. Whichever way we look at this, this outcome would be an instrumental success in policy terms, with a measurable impact. Yet it also invites an appraisal of individual behaviour and relies on the assumption that by matching savings the state is playing a legitimate role in shaping an individual’s outlook on the world. This is typical of a latent normativity in ABW research that does not rest upon – and cannot reasonably be said to assume – any explicit interpretation of ‘citizenship’.

This latent normativity is especially strong in the language of an ‘asset-effect’ and the assumptions we need to make if this is to be an analytical category with meaningful content. Thus, Michael Sherraden, arguably the founding father of ABW as a social policy movement, tells us that: ‘income only maintains consumption, but assets change the way people think and interact with the world’ (Sherraden, 1991: 6). This kind of thinking was also embedded in the official policy framework of the previous Labour government in Britain, with the Treasury asserting an ‘independent effect’ on life-chances, regardless of other factors such as class and education (HM Treasury, 2001). For the Institute for Public Policy Research (IPPR), one of the leading advocates of ABW, ‘having the opportunity to invest one’s own financial asset in a house, learning, or a business start-up fosters a sense of autonomy and responsibility’. And, ‘perhaps most significantly, there is an emerging body of evidence … which sets out the independent effect of asset-holding on key life-chances over and above the impact of income levels’ (Kelly and Lissauer, 2000: 8).

The IPPR embrace this theme, with perhaps a little more circumspection, in later work. ‘Asset-based welfare is based on the idea that financial assets bring positive benefits above and beyond simply allowing people to put off spending today in order to consume in the future, or to earn interest on investments’ (Sodha and Lister, 2006: 2, emphasis added). Here we are offered some plausible scenarios in which assets could play a very positive role in enhancing an individual’s well-being; for instance, allowing them to pursue further education, or giving them the financial freedom to leave an unhealthy relationship (Lister and Sodya, 2006: 2). More recently, McKnight and Karagiannaki (2013) have tentatively suggested that there may be an asset-effect that accounts for a greater individual sense of physical and mental well-being.

These authors, to differing degrees, are aware of the inherent risk of claiming a special ‘effect’ that could be explained by other factors, such as the ability to use assets to buy goods and services. There is also a recognition – again, to varying degrees – of counterfactual scenarios, in which the goods secured by wealth and assets in Britain could in principle be provided by stronger public services and a more generous benefits system (McKnight, 2011: 54). Nevertheless, even in the most tentative treatment of the issue we are still offered explanations of a possible independent effect. Thus, McKnight and Karagiannaki suggest that the association between asset-ownership early on in life and improved health outcomes later is due to the ability to access other welfare goods:

Wealth could allow an individual/family to have a higher standard of living through a healthier lifestyle in terms of living environment, diet, access to sporting facilities, holidays etc., in addition to access to medical and health services. The sense of well-being associated with a favourable asset-position could also influence the way individuals feel about their health and how they cope with any health problems. (McKnight and Karagiannaki, 2013: 138)

This is doubtless a sensible conclusion. What we have not been offered, however, is a convincing case that these effects are ‘intrinsic’ rather than ‘instrumental’ and achievable by other means. At this stage we might therefore conclude that the search for a pure ‘effect’ is simply misguided. After all, there is a strong instrumental case for ABW, without our having to find something ‘over and above’ this. Yet there is still a powerful intuition behind the desire to distribute assets more widely, and it is still possible that such a distribution might have a pure ‘effect’ on individuals. We are, after all, still in the early stages of this important debate. First, however, we need greater conceptual clarity in defining what we are looking for. In the following section I consider in greater detail the characteristics of an intrinsic effect. This leads us towards the second distinction – between existential identity effects and social identity effects – mapped out in the ‘Conceptual journey’ section.

What makes assets ‘special’? Identity and uniqueness

In order to find a special asset-effect we need to know what we are looking for. Our aim is to find an intrinsic asset-effect, or least to set the criteria for success. There are two ways of thinking about what an intrinsic effect would look like. The first is that the asset-effect is intrinsic in so far as it is unique – an effect that can only be achieved by asset-ownership. This is a very demanding standard, and is unlikely to ever be identified. But we can still make sense of it in terms of the distinction with instrumental goods: some goods are very clearly available, in principle, without the need for greater asset-distribution. A reliable and generous adult education system, funded out of general taxation, stands out as an example of a good that has no intrinsic connection with asset-ownership.

The distinction also serves a useful analytical purpose in helping us to think about a sliding scale, where a unique asset-effect may be very rare (or not exist at all) but where, nevertheless, we are able to do justice to the intuition that there are some things that asset-ownership is in fact likely to achieve better than other means. A sense of control over one’s home environment is an obvious example (though there are limits to ownership rights, and radically different rental rights could also give a far greater sense of control). The ability to build a business is another, though this can be achieved through debt finance, and relies on the (reasonable) assumption that the process of building the business – the sense of enterprise and achievement – counts as much as the riches it might bring. Yet it is still hard to think of pure cases where asset-ownership is the unique route to a particular end. The idea of intergenerational continuity of concern – through bequests – is perhaps the closest that we can get to this type of intrinsic effect, defined as a unique means of achieving certain ends.

But there is another dimension to the idea of an intrinsic asset-effect when we count personal and psychological transformation as one of those ends. Thus, the second way in which we can think about an ‘intrinsic’ asset-effect is through the identity effect introduced in the ‘Conceptual journey’ section. As we have seen, this identity effect can, in turn, be divided into two further categories; a ‘social identity effect’ and an ‘existential effect’. I shall address this second effect first.

We find strong elements of the idea of an existential effect in the greatest traditions of European philosophy, especially in Locke, Kant and Hegel (Locke, 1962; Kant, 1998; Hegel, 1991). In this tradition there is indeed something unique about property – its ability to bring about an existential transformation in the individual. Property-ownership, in this tradition, therefore has both the features we have attributed to a potential intrinsic ‘effect’; it is a unique means to an end (only achievable through asset-ownership), and the end itself is transformative for the individual. For both Kant and Hegel, the ability to take part of the external world and lay a claim on it is a necessary developmental process. In Hegel’s terms, it is the process by which the ‘will’ becomes ‘actual’. Ownership, crucially, is not an instrumental means to other ends.

Another crucial distinction – between what asset-ownership can do for you and what it does to you – helps make sense of this. This distinction cannot answer all our questions about the nature of the putative asset-effect, but it at least helps to elucidate the nature of the challenge faced by those seeking an asset-effect. We have thus taken a large step closer to what might be considered ‘special’ about asset-ownership. This last distinction leaves open the theoretical possibility that what an asset does to you (or has an effect on ‘who we are’) can be achieved by other means. But the distinction between ends and means (or the intrinsic and instrumental) is now so fine that, for our purposes, the concern remains largely theoretical. This is not to dismiss outright philosophical complication. For example, it is reasonable to ascribe a certain type of philosophical regression to the ‘for you/to you’ distinction, as we could reopen the intrinsic/instrumental distinction by suggesting that asset-ownership is the instrumental means to the end of psychological transformation (what the asset does to you). But we have nevertheless taken a large step closer to what might be considered ‘special’ about asset-ownership. The philosophical tradition stemming from Locke, Kant and Hegel opens up the bigger question of how ownership affects who we are, rather than what we do (or can do) with assets.

As we have just seen, historically this picture has had something of a metaphysical quality about it. A modern approach would be couched instead in the language of behavioural science. But this does not radically alter the conceptual and empirical challenges of finding a pure asset-effect: we still need to isolate an ownership effect that is unique, and the best place to look for this is in what it does to the individual, rather than what it does for her. Asset-ownership needs to be the unique route to a certain type of psychological and behavioural transformation. Moreover, just at the point where we have got closer to the idea of an asset-effect, we are inevitably pulled back into the question of what we want from it; how we might want ownership to change the beliefs of others, about both themselves and their place in society.

A final point is in order before we turn to the last conceptual category, the idea of a ‘social identity effect’. The same philosophical tradition I have just described is strongly associated with the role of property in maintaining a stable social order. And the idea that homeownership, in particular, gives individuals and families a ‘stake’ in society is still felt strongly in current conservative thinking and policy discourse. My account is therefore open to the objection that it misconstrues the real normative value of property in conservative thought. There is insufficient space to pursue this directly in this paper. Two immediate observations, however, stand out. Firstly, it is empirically questionable that ownership itself is what gives a household a stake in society; stable occupation in a quality social rented sector, in a functional and popular neighbourhood, could bring the same (putative and contestable) social gains. Secondly, those who do not have an ownership stake are tacitly assumed to have less of a commitment to society, and this takes us back to the relationship between property and citizenship, and, in turn, back to our core theme – the transformative potential of property.

The argument of the next section is that, when combined with this kind of property-owning political ideology, the current asset-effect evidence base leads in a distinctive political and policy direction, one that is radically at odds with the ‘progressive’ politics typically associated with asset-based welfare. A key point here is that the absence of hard evidence is far less problematic for advocates of what I refer to as ‘ownership as a duty’: there is still a duty to own even if there is no positive asset-effect on the individual. This is in contrast to accounts of ownership as a right of social inclusion, which do rely on such a positive effect. I turn to this right of social inclusion in the section following that below.

Asset-ownership as a duty of citizenship

The conservative language of ‘property-owning democracy’ has recently been co-opted by egalitarian minded political theorists (O’Neill and Williamson, 2012), who argue that greater distribution of property must augment what John Rawls has called ‘welfare state capitalism’ (Rawls, 2001). For our purposes, however, the important narrative is the tradition of property-owning democracy that has dominated politics and social policy in Britain at least since 1980. Unsurprisingly, this vision of property-owning democracy is presented in positive terms, as a great opportunity for individuals and as the foundation of a thriving society. But, crucially, it does not have to be positive for the individual; the ideal of independence and self-sufficiency is as much a duty of citizenship as it is a right. This means that the absence of an asset-effect, either instrumental or intrinsic, whilst politically important, need not cause a radical rethinking of either policy or politics. There will still be an expectation that we are to build our assets and to finance our own welfare consumption through divestment, and in a way that ties in with the ideology of the ‘rolling back’ of the state.

Conversely, a continued stream of evidence suggesting that there are such effects is something of an ideological and political gift to this account of asset-ownership and citizenship. Of course, the search for such evidence is laudable. No factual or intellectual enquiry should be restricted out of fear of the results, whatever one’s political persuasion. But a note of caution does need to be sounded in this case: if the intellectual desire to find an asset-effect leads to the kind of elisions I have described (especially the elision of instrumental and intrinsic value), the evidence is more open to spurious interpretation and unfounded political rhetoric.

A comparison with two other ‘effect’ debates helps draw this out. The concept of a ‘tenure effect’ is posited as an explanation for the disproportionate number of workless households in social housing in the UK (roughly 40 per cent of all working age tenants), controlling for the fact that workless households are more likely to qualify for social housing in the first place. As observed by John Hills, ‘for any given number of overlapping disadvantages, those in social housing have lower employment rates … Even controlling for a very wide range of personal characteristics, the likelihood of someone in social housing being employed appears significantly lower than those in other tenures’ (Hills, 2007: 102, 111). Here, then, we are looking to identify a mechanism that does something to an individual’s identity or behaviour.

One possible explanation is that there is a ‘neighbourhood effect’, suggesting that it is the high concentration of social housing in certain types of deprived area that has a negative impact on individuals and their life-chances (e.g. Musterd and Andersson, 2006; Green and White, 2007), rather than social housing per se. This is not simply the claim that disadvantage is aggregated, or that the neighbourhood is deprived by relatively simple causal mechanisms, such as disconnection from appropriate labour markets. Rather, the claim is that there is something ‘over and above’ this that is entrenching disadvantage. If this is the case, we are likely to find that the location and density of social housing has some form of ‘social identity effect’, with individual expectations lowered and horizons limited; a case of adaptive preferences and ‘negative’ social capital. In part this could be because of a sense of shame and stigma created by separation from mainstream society, and fuelled by a virulent ‘anti-scrounger’ media. Notably, this looks very much like the ‘social identity effect’ that has been mapped out in the ‘Conceptual journey’ section. If this is the case, there is no intrinsic disvalue in social renting. Rather, it is a contingent feature of the way it is distributed. Asset-ownership is playing a role in shaping these identities, but the mechanism is part of a negative process in which those who don’t own are judged to be inferior.

But there is also a competing interpretation, and one that relies on the same body of evidence, using different normative assumptions. This explanation claims an intrinsic effect: it is the tenure itself – the type of housing an individual lives in – that changes identity and behaviour. On this interpretation, social housing slots easily into the counterfactual of the ‘intrinsic effect’ case for homeownership: whereas the ‘effect’ of ownership is to transform the individual into a responsible and ‘independent’ citizen, the effect of social housing is precisely the opposite. Hence, those born into social housing do worse because of the welfare state and ‘the appalling incentives that social tenants face’ (Morton, 2010: 5). Here we see something closer to what we have described as a pure or ‘existential’ identity effect. If ownership creates a new identity of independence (over and above financial and practical freedoms), the opposite identity and behaviour will inevitably be found in social housing. Whilst we have touched upon these claims in the ‘Owing-up’ section, it is interesting to note here the historically persistent tendency to universalise this type of identity effect: ‘[t]hough the effects are getting worse, social housing has always damaged equality of opportunity’ (Morton, 2010: 52, emphasis added). ¹

There are, then, two types of explanation for the persistently high unemployment in the UK’s social housing, both relying on the concept of an ‘effect’, and both with very different ideological and political implications. Whereas a neighbourhood effect based on disconnection and poor social capital naturally leads to a ‘social inclusion’ framework most often associated with ‘progressive’ policy, contemporary interpretations of a tenure effect have been closely associated with a rolling back of the welfare state: social housing causes dependency, so remove the cause.

The broader point here is that both explanations have started from the same basic problem – high levels of worklessness in social housing – and, crucially, the same evidence base (Hills, 2007; Feinstein et al., 2008). And in fact, just as in the case of an asset-effect, we simply do not know as yet if there is indeed either a ‘tenure’ or a ‘neighbourhood’ effect. The evidence is not there. But this has not stopped the idea of an effect being co-opted by the ideology (and policy) of citizenship, independence and duty. This, I suggest, has been made possible because insufficient attention has been paid to the normative dimensions of the putative neighbourhood and tenure ‘effects’ in the ‘academic’ debate, leaving a vacuum that is easily filled by more ‘political’ interpretations of the evidence. Likewise, as we have seen, the competing ideology of ABW as independence is ready to fill any normative vacuum left behind by straight empirical fact, just as narratives of self-sufficiency and independence have come to dominate the debate about social housing and worklessness.

Asset-ownership as a right of social inclusion

The absence of a true asset-effect therefore, in some respects, leaves progressive advocates of ABW in a more difficult position than the advocates of a self-sufficiency and small-state agenda. No clear normative position has been articulated by the ‘progressives’ and the progressive case has often been based on the questionable assumption that there is something ‘unique’ about asset-ownership.

One option is to retreat to the core purpose of ABW as an instrumental tool of social policy. There are very good reasons in favour of this social policy framework, giving households a greater ability to withstand sudden income shocks, for example. But there is a normative loss as well, as it seems that there is no longer anything ‘special’ about assets, over and above income or services that lead (or could lead) to equally good outcomes for individuals. Of course, there are still strong normative reasons for objecting to great inequalities of wealth. At the top of the list we might place the ideal of desert, arguing that the luck of the position we are born into – combined with the huge range of public goods that allow us to pursue our plans and aspirations – renders implausible claims that vast piles of private wealth are exclusively the result of individual endeavour (Murphy and Nagel, 2002).

Yet this type of argument leads us in a different direction than a more positive account of the intrinsic value of asset-ownership would. The upshot of the desert account is that it would seek to be restrictive, redrawing the boundary of individual property rights, claiming back for the public realm goods previously characterised as private. What is attractive about the claim to the intrinsic value of asset-ownership, however, is that it is expansionary: if asset-ownership is intrinsically valuable as a developmental human good there is a powerful case for promoting it as a right. This could, we should note, apply without a concomitant restriction of the rights of the richer members of society. Indeed, this feature of the (putative) intrinsic effect helps explain the natural ease with which the ABW agenda, originating from the left of the political spectrum in Britain (Kelly and Lissauer, 2000; Nissan and Le Grand, 2000; Bynner and Paxton, 2001), has been reclaimed by those further to the right of the spectrum (Wind-Cowie, 2009; Blonde, 2010). Whilst this co-option has thus far proceeded openly and in good faith, the ‘asset-effect’ is still vulnerable to political exploitation, with a duty of self-sufficiency being sold as a positive right.

There is, however, an alternative account of citizenship and asset-ownership, one that need not rely on the empirical identification of an intrinsic asset-effect. Clearly, there needs to be a recognition of the instrumental potential of ABW. But the crucial normative dimension is to be found in a more contingent account of the social status of property in contemporary society: it is not that ABW is to create better citizens, but that a full sense of citizenship requires access to the goods that come with asset-ownership. This is a consequence of the way in which asset-ownership has an important ‘social identity effect’. Given the enormous social and cultural resonance that ownership has in British society, the justification for wealth and distributing assets through ABW could be simply based on the discourse of social exclusion: the standards and norms of the day require significant asset-holdings if the individual is to feel part of society, so progressive social policy should facilitate that participation.

Note, however, that advocacy of this stance does not require one to make the further claim that asset-ownership has a positive transformative effect on the individual, so there is no ‘existential’ or behavioural claim here. The transformative effect would be the process of crossing the barrier of exclusion; if ownership were not so culturally important, being excluded would not have the same psychological and social impact. One can therefore advocate ABW whilst remaining agnostic on the issue of whether asset-ownership and market activity have any further positive effect on well-being, over and above the fact that the exclusion itself is likely to have a detrimental effect on an individual’s well-being. One might feel excluded from a national celebration of Frisbees if one doesn’t own one, but that does not amount to a coherent case for the intrinsic value of Frisbees.

Conclusion

Private property has always been a defining characteristic of liberal democracies. Historically, it has been a requirement of full citizenship – a precondition of the right to vote. ‘Ownership’ has always meant more than the possession of material goods. Of course, these material goods can be of great practical utility, and we rightly worry about inequalities of wealth on this basis alone. Unequal wealth almost inevitably – though not necessarily – entails significant inequalities in both current standards of living and future ‘life-chances’. Social policy research should be actively engaged in understanding the extent of these inequalities, as well as the extent to which they can be compensated for by public services and income transfers.

Almost certainly, this process will come to the conclusion that some form of ‘asset-based welfare’ is a desirable addition to more traditional welfare provision. But there is significantly more at stake here; the depth of normativity is greater, and more complex, than even the ‘citizenship’ strand that the debate currently acknowledges. Until we begin to work through this complexity there can be little clarity of purpose in the search for an asset-effect. There can be still less clarity regarding the desirability of such an ‘effect’ – and no real debate on what we really want from asset-based welfare. For those who assume that the search for an asset-effect leads in an egalitarian or ‘progressive’ direction, the simple message is ‘proceed with caution’: it is too early to lay an exclusive normative or political claim on the value and implications of the putative ‘asset-effect’.