BARRIERS TO EMPLOYMENT AMONG LONGTERM BENEFICIARIES A REVIEW OF

Internationally and domestically, there has been increased policy focus on reducing barriers to employment among working-age beneficiaries. Recent reforms in the United Kingdom (UK) and the United States (US), for example, were aimed at both increasing motivation for work (by addressing financial disincentives and instituting sanctions) and providing services needed to support employment. In New Zealand, as elsewhere, there is now growing attention to the employment prospects of long-term beneficiaries, who may face multiple labour market disadvantages. A review of what is known about the special needs of those who are most likely to be on benefits for a long period of time will inform our understanding of potential policy options.

The purpose of this report is to document the recent international literature on the topic of potential barriers to employment among working-age beneficiaries, with a particular focus on long-term beneficiaries.^¹ The objective of this literature review is to provide an overview of key findings, in order to provide a solid evidence base for policy analysis. The review aims to answer the following questions:

What are the key barriers to employment for beneficiaries?
Which employment barriers apply particularly to long-term beneficiaries?
How do these barriers vary by benefit type?
How does the international literature on barriers interface with what we know about the barrier profile of long-term beneficiaries in New Zealand?

In order to answer these questions, the literature review focuses on recent policy-relevant work in the UK, Australia, the US, Canada and New Zealand up to 2002. The extent of the literature on the topic required careful targeting of sources most likely to provide timely, recent, high-quality and policy-relevant research results and review articles. Key academic and research institutes and government agencies that have working-paper and discussion-paper series and publications that are available via the web served as the key sources. The Appendix lists these websites and describes the methods used to construct the literature review. Generally, the research papers reviewed were aimed at identifying employment barriers among the benefit population, and understanding the prevalence, nature and significance of the various barriers.^²

The literature reviewed here provided a thorough survey of the barriers to employment that have been identified by and are now being discussed at the policy research centres consulted. However, the papers cited here do not represent an exhaustive list of the evidence for each particular barrier. In cases of obvious gaps, a review of academic literature was carried out and summarised in this report.

The review consists of three parts. First, the report identifies, describes and documents evidence for a host of employment barriers for all beneficiary types at the personal, family, social/community, and institutional levels. Second, several key issues for the long-term beneficiary population specifically are discussed. Finally, the report summarises the knowledge base in New Zealand regarding long-term beneficiaries and their employment barriers.

2. Barriers to employment

Below is a summary of findings on barriers to employment among beneficiaries that have been identified in the international literature, including that of New Zealand. Although the barriers are listed individually, it is important to remember that a key finding from this review is that long-term beneficiaries face multiple and intersecting barriers. We return to this point repeatedly in the discussion on key issues for long-term beneficiaries.

The barriers are grouped by the level at which they occur – personal, family, social/community, work-related and benefit-system. For the most part, barriers included in the personal, family and social/community sections are not specifically work-related but do have an impact on employment. The barriers listed in the work-related section refer to both personal and structural barriers, as well as their interaction.

2.1 Personal Barriers

2.1.1 Poor health and disability

Poor health and disability are significant barriers to employment and may affect larger proportions of the unemployed and lone-parent beneficiary populations than previously recognised. In some contexts (eg the UK), poor health and childcare are the two most significant barriers. Poor health is one of a host of factors associated with economic disadvantage.

The basic evidence

Work-limiting health problems and disabilities have been identified as significant barriers to employment among all beneficiary types, including lone parents (Baker 2002; Baker and Tippin 2003; Curtis 2001; Fraker and Prindle 1996; Hales et al 2000; Marsh et al 2001; McKay 2002; Olson and Pavetti 1996; Parker 1997; Polit, London and Martinez 2001; Riccio and Freedman 1995; Stephenson 2001), the unemployed (Marsh et al 2001; McKay et al 1999; Trickey et al 1998), non-employed couples (Dorsett 2001; Marsh et al 2001; McKay et al 1999; Stephenson 2001), and, of course, people with disabilities (Department for Work and Pensions 2003; Jonczyk and Smith 1990; Loumidis et al 2001).

Prevalence

Not surprisingly, medical conditions that limit everyday activities are nearly universal among people with disabilities: in the UK, 98% of incapacity beneficiaries report a medical condition, with the percentage declining to 80% among those deemed closest to the labour market (Loumidis et al 2001). Among all long-term beneficiaries in the UK, three-quarters fall into the “sick or disabled” category (Department for Work and Pensions 2003).

Among beneficiary types other than the sick and disabled, disabilities may play a larger role in non-employment than is sometimes recognised. According to 1990 nationally representative US data, 16–20% of female welfare recipients had a work-limiting disability, and over half of these women were limited in their ability to perform basic functions, such as walking, bathing and eating (Loprest and Acs 1995). More recently, from a four-city study in the US, Polit, London and Martinez (2001) found that 44% of non-working welfare recipients had a physical condition that limited moderate activity. Ethnographic interviews from the study also found serious health problems among women who had reported in the survey that they were in good health, suggesting that under-reporting may be a problem.

In a recent New Zealand study, recipients of the Domestic Purposes Benefit (DPB) reported poorer health than did women in the general population (Baker 2002). In the UK, one-third of all income support claimants have a long-term health problem and/or caring responsibilities (Gardiner 1997). In a review of research on non-employed couples, Millar and Ridge (2001) describe the most common scenario as one in which the male partner has an illness or disability and the female partner has caring responsibilities for her partner, their children or both. In addition, a significant minority of the female partners reports health problems (20%, compared with 69% who report not working for family reasons) (McKay et al 1999).

Nature and significance of the barrier

Of the barriers reviewed here, health problems and disabilities are particularly significant in inhibiting labour force attachment among beneficiaries. Lone parents and unemployed couples with long-standing health problems are less likely to move into employment. Among non-employed lone parents in a UK study, a move into employment was 95% more likely among those not reporting a long-standing health problem, and non-employed couples showed a similar pattern (McKay 2002). Over one-quarter of DPB recipients surveyed in New Zealand said that their own poor health prevented them from taking up paid employment (Baker 2002). The manner in which health affects their employment is multi-dimensional and ranges from serious physical diseases of their own, chronic conditions such as asthma among their children, and feelings about being unable to cope emotionally with both a health condition and employment (Baker and Tippin 2003).

In a study of a Californian welfare-to-work programme, Riccio and Freedman (1995) report that serious health problems inhibited continuous employment among a substantial minority of programme participants. There was a higher prevalence among long-term beneficiaries, 30% of whom received a deferral from participation for a medically verified illness. In one of Iowa’s welfare-to-work programs, one-third of non-compliant recipients who were cut from the rolls reported a serious personal or health problem, which suggests that such problems inhibited not only employment but also programme participation. Among unemployed beneficiaries in the UK who had moved to the Jobseekers’ Allowance (JSA) from the Incapacity Benefit (IB), half were still on JSA 5–10 months later, and almost 70% said that their health had not improved since leaving the IB (Ashworth, Hartfree and Stephenson 2001).

Severity

Both the severity and perceived permanence of the medical condition affect the extent to which it acts as a barrier to employment. In the UK, Stephenson (2001) found that the perceived permanence of a condition can make a health problem a “fairly intractable barrier”, contributing to a pessimistic mood among beneficiaries in terms of their ability to work. Among incapacity beneficiaries deemed closer to the labour market, the severity of their medical condition affected their expectations about being able to work (Loumidis et al 2001). Similarly, Berthoud, Lakey and McKay (1993) found a linear negative relationship between impairment severity and employment among people with disabilities in the UK. However, severity interacts with age, and lower severity scores have a greater effect on non-employment at older ages. This may be because older people with disabilities face additional barriers that depress employment even at lower levels of impairment severity. Similarly, among people with disabilities in Australia, Jonczyk and Smith (1990) found that the relative importance of the medical condition as a barrier to employment declined with age. For older clients with disabilities, age discrimination, employer attitudes toward people with disabilities, and atrophied skills played a larger role than the medical condition alone.

Association with other barriers

Long-standing illness is associated with a host of factors that may be clustered among certain individuals “at risk” of being or becoming long-term beneficiaries. For example, Finlayson et al (2000) found that, among a cohort of lone parents in the UK that was followed between 1991 and 1998, age, employment status, smoking, severe economic hardship and history of domestic violence were all associated with having a long-term illness that inhibited employment in 1998. The material hardship faced by beneficiaries makes it difficult for many to pay health-related costs, such as doctor’s visits, so health may suffer even more (Duncan, Kerekere and Malaulau 1996). Using Canada’s National Population Health Survey, Curtis (2001) found that lone mothers have lower unconditional health status than married mothers, but that the differences disappear once age, income, education, lifestyle factors and family size are taken into account. Baker and Tippin (2003) argue that, in order to understand the health–work nexus, “we must consider how poverty dominates the lives of [lone-parent beneficiaries] making them more vulnerable to health-related setbacks than more advantaged members of society” (p. 23). Thus, while poor health may be a critical indicator of potential current employment probabilities, it is important to recognise that its underlying cause may be rooted in a range of other problems or disadvantages that may have accumulated across the life course (see Section 3.4).

2.1.2 Mental illness and psychological distress

There is a well-documented association between unemployment and poor mental health but the direction of causation is more complex than among the lone-parent population. Forms of psychological distress (such as low morale, low self-esteem and lack of confidence) may affect a significant proportion of all beneficiaries. Long-term recipients are especially likely to have a psychiatric disorder (such as clinical depression) and may be at greater risk of psychological distress. Contributing contextual factors, such as material hardship, need to be taken into account. In addition, mental illness often coexists with substance use disorders. Clinical depression and other psychiatric disorders are significant barriers among a substantial minority of lone mothers in the US.

The basic evidence

Several studies in the US have documented an association between low socio-economic status and mental illness (see Derr, Douglas and Pavetti 2001 for review), while others have documented higher rates of mental illness and depressive symptoms among beneficiaries compared with the general population (Arthur et al 1999; Barusch, Taylor and Abu-Bader 1999; Brooks and Buckner 1996; Coiro 2001; Danziger et al 2000; Loumidis et al 2001; Marcenko and Fagan 1996; Olson and Pavetti 1996; Sweeney 2000; Zedlewski 1999). More generally, psychological distress, as indicated by low self-esteem, low morale and lack of confidence, is common and has been found in several countries to inhibit employment (Barnes, Thornton and Campbell 1998; Dawson, Dickens and Finer 2000; Finch et al 1999; Finlayson and Marsh 1998; Lewis et al 2000). Most research on the association between unemployment and poor mental health has been conducted in Nordic countries (see Bjorklund and Eriksson 1998 for review), although the association has also been documented in the countries covered in this review (eg Breslin and Mustard 2003 for Canada; Comino et al 2003 for Australia; Pernice and Long 1996 for New Zealand). Overall, the research suggests that the unemployed have poorer mental health than do others, and longitudinal research suggests that unemployment contributes to deteriorating mental health. Thus, while mental health problems may not “cause” unemployment, deteriorating mental health during unemployment may become an increasingly important barrier to employment. In addition, mental health problems may coexist with substance abuse problems (see Section 2.1.4).

Prevalence

Mental illness is subsumed under “disability” for many affected individuals who face difficulties with employment because of their condition. In a study of disabled beneficiaries participating in the Personal Adviser Pilot in the UK, one-third of clients reported a mental health condition as their main health problem (Arthur et al 1999). Similarly, Loumidis et al (2001) report mental health problems as the main medical condition among incapacity beneficiaries judged to be closer to the labour market, affecting 29% of the sample. In the US, Sweeney (2000) estimates that between one-fourth and one-third of welfare recipients have a serious mental illness that could interfere with their ability to find and keep employment. Major depression seems to be the most common disorder (Danziger et al 2000). In a study of long-term welfare recipients in Utah, Barusch, Taylor and Abu-Bader (1999) found that 42% had clinical depression in the year before the interview, a rate seven times higher than among the general population.

Less research has quantified the extent of more general psychological distress experienced by beneficiaries. In a study of UK lone parents, Finlayson and Marsh (1998) found that almost one-quarter felt that life’s problems were too much for them and one-half felt, at times, that they were useless. Almost one-fifth felt that they were failures. In a study of low-income families in the UK, Marsh et al (2001) found that 12% of females in couples and 14% of male partners felt that they were failures; 18% and 12%, respectively, said that they did not have a positive attitude about themselves.

Nature and significance of the barrier

The higher rates of mental illness among beneficiaries compared with the general population clearly indicate that this is a barrier to employment (Barusch, Taylor and Abu-Bader 1999; Sweeney 2001), although the direction of causation is difficult to determine. In the US, Coiro (2001) found that the more depressive symptoms a mother on welfare has, the less likely she is to leave welfare over a two-year period. Also in the US, Danziger and Seefeldt (2000) found that mental health problems differentiate the working from non-employed beneficiaries. In contrast, Brooks and Buckner (1996), while finding high rates of mental illness in their sample of homeless and housed low-income women, did not find that the state of mental health determined who worked and who did not. However, theirs was a small and especially disadvantaged sample.

The relationship between unemployment and mental illness is complex. Generally, longitudinal research suggests that becoming unemployed leads to an increase in psychological distress and depression (Bjorklund and Eriksson 1998; Harquist and Starrin 1996; Janlert 1997), while re-employment leads to recovery or improved mental health (Claussen 1999; Pernice and Long 1996). One Canadian study suggests that this is the case only for older (31–55-year-old) workers (Breslin and Mustard 2003). Once mental health problems take hold, they can then act as a barrier to further employment. In addition, there is a selection process going on such that those with mental health problems will have a harder time finding and keeping employment and will thus be overrepresented amongst the long-term unemployed (Janlert 1997).

In addition to mental illness, less severe but perhaps more common markers of psychological distress (such as low self-esteem, and lack of confidence and morale) are frequently cited as contributing to, and possibly being caused by, long-term benefit receipt. A lack of confidence can affect a beneficiary’s assessment of their employability and can cause them to worry about how they might come across in interviews. In a UK study on disability, Barnes, Thornton and Campbell (1998) report that a lack of confidence has been identified as a barrier to employment by people with disabilities themselves and by a number of researchers studying the issue. Three UK qualitative studies identify lack of self-confidence, lack of self-esteem and low morale as barriers to employment among lone parents (Dawson, Dickens and Finer 2000; Finch et al 1999; Lewis et al 2000).

Evidence that these psychological characteristics inhibit employment among beneficiaries comes from evaluations of programmes that have addressed these issues either directly or indirectly. In a study of lone parents participating in the UK’s Personal Adviser scheme, Hales et al (2000) found that many lone parents who moved into work identified the boost in confidence that their Personal Adviser gave them as a major factor in their success. Similarly, in a study of “innovative schemes” in the UK aimed at moving lone parents into employment, the most successful approaches worked by boosting confidence and motivation and broadening horizons (Woodfield and Finch 1999). Confidence building coupled with social support were key elements.

Stress

Stress may play a key role in high levels of employment-limiting psychological distress among beneficiaries. In the US, Kalil et al (1998) found that first-time welfare recipients experienced an average of eight stressful life events in a year, compared with a general population average of one or two. Lone-mother beneficiaries in North Carolina who experienced parenting stress, caused by low levels of personal control and high levels of problem child behaviour, were less likely to complete an education and training course (Orthner and Neenan 1996). Parker (1994) found that improved levels of personal control (a key element of stress) were associated with reduced welfare use. The cumulative effect of stress over time can weaken a person’s sense of personal control, confidence, self-mastery and efficacy, all of which can then have negative impacts on employment probability (Lennon 2001).

Reciprocal causation

Low morale, low levels of confidence, low self-esteem and depression can all inhibit job search and sustainability among beneficiaries. Likewise, living on a benefit and not being employed can contribute to low morale, low levels of confidence, low self-esteem and more serious forms of mental illness. Thus, causation can run both ways and through various mechanisms. Danziger, Carlson and Henly (2001) found that, among men and women who had lost General Assistance benefits in the last year, those who had worked in that year were less likely to be depressed and more likely to be satisfied with life.

Living with material hardship, especially over the long term, contributes to low morale and depression (Finlayson et al 2000; Marsh et al 2001; Scott et al 1999). Benefit levels in the countries covered in this review are typically not high enough to guarantee that a beneficiary’s life will be without material hardship. In Sweden, the more generous income-replacement benefit scheme has been found to be associated with improved mental health among the unemployed in contrast to less generous cash benefits (Strandh 2001).

Receiving means-tested benefits, as opposed to entitlements, may also contribute to poor mental health, above and beyond any effects on material hardship. Two studies found that, even controlling for a range of associated factors including household income, unemployed men and women who received means-tested benefits or no benefits had significantly higher levels of depression and ill-health than the full-time employed. In contrast, the unemployed who received entitlements had similar levels of depression compared with the full-time employed (Rodriguez, Frongillo and Chandra 2001; Rodriguez, Lasch and Mead 1997). Lack of attachment to the labour market may also contribute to feelings of isolation and decreased confidence levels.

There is evidence that the longer a person lives in these conditions (with material hardship and away from employment), the more entrenched the problems, like a lack of confidence, can become, and the more they may act as barriers to employment (Coiro 2001). In the aggregate, the effects can then compound over time: individuals with higher morale and less hardship move into work more quickly, while those with greater hardship and lower morale remain on a benefit longer and then become further discouraged and mired in economic hardship (Millar and Ridge 2001). As Finlayson and Marsh (1998) point out, lone parents who are characterised by a number of disadvantages (younger, never married, poorly qualified, live in poor neighbourhoods) are more likely to have low self-esteem and morale, to have more difficulty finding employment and, over time, to “see their position in the labour market get weaker” (p. 2).

However, employment in and of itself should not be viewed as a panacea for psychological distress. Improvements in financial circumstances, as might accompany a move into a well-paying job, might lead to higher levels of feelings of self-mastery among lone parents (Ford, Marsh, and Finlayson 1998), thus increasing their chances of remaining employed. However, in a US study, lone mothers working in low-wage jobs had similar levels of depressive symptoms and feelings of hopelessness as did lone mothers still on welfare (Pettersen and Friel 2001). This supports the idea that material hardship may be a root “cause” of psychological distress. Similarly, Ahluwalia et al (2001) found that participation in welfare-to-work schemes by those who already had depressive symptoms sometimes exacerbated these symptoms. While the cause is not clear, evidence suggests that the increase in depressive symptoms might be caused by the negative impact on children of the mother’s participation in the work scheme. Programme participants in one welfare-to-work scheme reported more depressive symptoms and lower levels of feelings of warmth toward their children compared with a control group of non-participants, and their children exhibited higher levels of behavioural problems.

In summary, mental illness and psychological distress can act as major barriers to employment and may have stronger cumulative effects over time. However, it is important to pay attention to the contextual factors, such as material hardship, that contribute to these conditions.

2.1.3 Learning disabilities

Few studies examine learning disabilities specifically, although a proportion of the general population of beneficiaries with disabilities is learning disabled. Limited evidence from the US for the lone-parent beneficiary population suggests that levels might be high and that learning disabilities often go undetected, creating a potentially “hidden” barrier.

Only a few of the studies covered here singled out learning disabilities as a significant barrier to employment. Presumably, some of the research aimed specifically at incapacity beneficiaries and people with disabilities generally covers the learning disabled, but they are typically not discussed as a separate group. Although few attempts have been made to assess the extent of learning disabilities among other beneficiary types, there is some limited evidence in the US that rates might be high among lone-mother beneficiaries. Summarising findings from three state studies, Sweeney (2000) reports that between one-fifth and one-third of welfare recipients have learning disabilities, with the rate in Washington State possibly as high as one-half. In a Kansas State pilot study, three-quarters of programme participants diagnosed with a learning disability had never before been identified as learning disabled, so many more might be affected than programme planners and caseworkers are aware. The high levels of learning disabilities found in these studies suggest that this is a factor that may require more systematic evaluation.

Learning disabilities could have serious effects on the employability of beneficiaries. However, as in the general population, they may go undetected. Undetected learning disabilities in childhood could affect subsequent educational attainment and work experience. Thus, some proportion of the low education levels and low work-experience levels among long-term beneficiaries might be attributable to learning disabilities. This is a topic for further research.

2.1.4 Substance abuse and dependence

Drug or alcohol dependence and misuse may affect a small proportion of lone mothers in the US (estimates vary widely) but is a significant barrier when present. There is a well-documented association between unemployment and substance use disorders, with unemployment increasing the risk of developing drug- and alcohol-related problems.

The basic evidence

The extent to which alcohol or drug use and dependence affects the employability of beneficiaries is an understudied area. Evidence from the US provides both a wide range of estimates of substance abuse (Sweeney 2000) and the finding that substance dependence is rare and not a significant barrier to employment overall (Danziger and Seefeldt 2000; Marcenko and Fagan 1996). Many studies have documented an association between alcohol misuse and unemployment (Dooley and Prause 1998; Khan, Murray and Barnes 2002; Montgomery et al 1998; Mallahy and Sindelar 1996; Terza 2002) but generally have not quantified prevalence among unemployment beneficiaries at a national level.

Prevalence

Prevalence rates among unemployed beneficiaries and disabled beneficiaries, and among lone-parent beneficiaries outside of the US, were not available in the studies covered in this review, most likely because such data is generally not available on a national level. Generally, it appears that, while use or abuse of alcohol or drugs may affect a sizeable minority of beneficiaries, dependency rates are very low. Sweeney (2000) reports that the range of estimates on the prevalence of substance abuse among welfare recipients in the US is 2–20%. In another study, Pollack et al (2002) found that about 20% of beneficiaries reported use of an illegal drug in the past year, but only a small proportion could be classified as drug- or alcohol-dependent. Other US studies suggest that substance dependency among welfare recipients is rare (Danziger and Seefeldt 2000).

Nature and significance of the barrier

Substance use disorders negatively affect employability (Bray et al 2000; Bryant et al 1996; Fergusson, Horwood and Lynskey 1997; Hall et al 1999), and may be a factor in long-term benefit receipt. Pollack et al (2002), using US nationally representative cross-sectional and Michigan panel data, report that welfare recipients appear to be more likely than non-recipients to use or be dependent on illicit drugs. Long-term beneficiaries made up a large proportion of beneficiaries who reported recent cocaine use. However, the proportion of long-term beneficiaries who reported cocaine use was still very small. Marcenko and Fagan (1996) conclude that, overall, substance dependence was not a significant barrier to employment among US welfare recipients.

Generally, longitudinal research in the US, Canada and the UK suggests that unemployment causes an increase in alcohol misuse (Dooley and Prause 1998; Khan, Murray and Barnes 2002; Montgomery et al 1998; Mallahy and Sindelar 1996; Terza 2002). Khan, Murray and Barnes (2002) find that unemployment decreases alcohol use and misuse in the short term but increases it in the long term. In addition, they find that increased levels of poverty also increase alcohol use and misuse. Similarly, Dooley and Prause (1998) find that job loss increases the risk of alcohol use and misuse but so does underemployment and being employed at poverty-level wages. Thus, “work first” efforts that are aimed at getting beneficiaries into jobs quickly may not contribute to employment sustainability among those at risk of developing alcohol problems unless they are connected to wages above the poverty level.

Finally, substance misuse problems often coexist with mental health problems (Drake and Wallach 2000), although the topic of barriers to employment among the “dually diagnosed” (those with both substance use and mental health disorders) is understudied (Laudet et al 2002). In one of the few studies to examine the issue, Laudet et al (2002) find that non-employed, dually diagnosed individuals expressed high interest in working but cited multiple barriers to employment. Mental health symptoms and perceived obstacles (such as fear of failure and insufficient skills) were significant barriers to employment, whereas substance use was not.

Providing cash benefits for beneficiaries whose main disability is a substance use disorder is controversial (Rosenheck 1997). However, longitudinal research by Swartz, Lurigio and Goldstein (2000) suggests that withdrawing such benefits (as in the US) may create a residual population that is too impaired to work because of severe substance use and psychiatric disorders. In follow-up interviews with former substance-use disability beneficiaries whose benefits were withdrawn, the authors found that those who remained off federal benefits but were either unemployed or underemployed were five times more likely than those who had found employment to be drug dependent and seven times more likely to have two or more psychiatric disorders.

2.1.5 Attitudes

A great deal of evidence suggests that beneficiaries generally want to work for personal (eg autonomy, respect) and financial reasons; thus, it is usually some other barrier or set of barriers keeping the non-employed from working. However, there is evidence that some lone mothers have a commitment to care for their child(ren) that takes precedence over their desire to work, and that a small proportion of beneficiaries may feel apathetic toward work. In addition, discouragement, and lack of motivation and confidence may be more significant as barriers among long-term beneficiaries.

The basic evidence

In general, studies in all countries find that beneficiaries (of all types) want to work and have positive attitudes toward work (Danziger, Carlson and Henly 2001; Department of Labour (DoL) and Ministry of Social Development (MSD) 2001; Duncan, Kerekere and Malaulau 1996; Finlayson and Marsh 1998; Hedges and Sykes 2001; Johnson and Fielding 2000; Jordan 1989; Kempson 1996; Kempson, Ford and England 1997; Marsh 1997; Noble, Smith and Cheung 1998; Puniard and Harrington 1993; Rank 1994; Stephenson 2001).

Prevalence

The studies cited above repeatedly report that the vast majority of beneficiaries and of low-income families generally holds positive attitudes toward work, and not only well-paying work. For example, in the study of low-income families in the UK by Marsh et al (2001), only about one-quarter of lone parents and partners in couples disagreed with the statement “Having almost any job is better than being unemployed”. However, apathy and discouragement may be higher among long-term beneficiaries, who are not usually the specific focus of these studies. Stephenson’s (2001) indepth interviews with 44 low-income families in the UK revealed that a very small number described being apathetic toward work, having “just got used to not working”. In New Zealand, Parker (1997) found high levels of discouragement among very long-term unemployed job seekers, with 63% identifying discouragement as a barrier to finding employment. Päkehä were more likely than other ethnic groups, and men more likely than women, to report discouragement as a barrier. In general, this is an understudied area. However, it is important to note that negative attitudes toward employment among long-term beneficiaries may partly reflect perceptions about employer discrimination (see Section 2.4.3).

Nature and significance of the barrier

Beneficiaries’ reasons for wanting to work are not only (or even most importantly) financial. Lone mothers interviewed in several studies in different countries identified respect (including self-respect), role-modeling for children, independence, social contact, satisfaction, absence of stigma and identity issues as important reasons for wanting to work (Puniard and Harrington 1993; Rank 1994; Scott et al 2001; Stephenson 2001). Unemployed couples and unemployed individuals with disabilities also mentioned relief from the boredom of staying home, reduced social isolation and increased social life (Hedges and Sykes 2001; Stephenson 2001).

Because the vast majority of beneficiaries in these studies express a desire to work, the researchers generally conclude that attitudes per se are not a major barrier for beneficiaries unable or apparently unwilling to find work. Instead, they suggest that other barriers prevent beneficiaries from moving closer to employment, including physical or mental illness (Jordan 1989), low or diminished skills (Hedges and Sykes 2001; Johnson and Fielding 2000) and real or perceived employer discrimination against people with disabilities (Hedges and Sykes 2001), older workers (Encel 2000) or welfare recipients (Crowell 2001).

The extent to which attitudes act as employment barriers varies according to beneficiary type and family role. In the case of older unemployed individuals and those with disabilities, experiences with and perceptions of employer discrimination, as well as unsuccessful job search, contribute to feelings of discouragement. In the case of older unemployed workers in New Zealand, Patterson (1999) suggests that many do face the attitudinal barrier of not wanting to seek government aid in finding employment, but this same barrier prevents them from going on a benefit as well.

Beneficiaries with young children are more likely to hold ambivalent views of work (Kempson, Ford and England 1997; Noble, Smith and Cheung 1998; Scott et al 2001). In a study aimed at understanding the impact of in-work benefits on beneficiaries’ incentives to work, Kempson, Ford and England (1997) also provide some interesting evidence regarding beneficiaries’ attitudes toward work generally. Attitudes seemed to follow respondents’ role in the family. Unemployed male breadwinners were especially committed to employment, and their strong desire to work overrode financial considerations. At the other end of the scale were lone mothers with young children, who were more committed to providing care for their child. Other than this section of the unemployed, however, beneficiaries generally had positive attitudes toward work and had modest wage aspirations. In New Zealand, a government report on DPB recipients indicated high levels of motivation to work, given suitable employment (DoL and MSD 2001).

2.1.6 Criminal convictions

Surprisingly little research attention has been paid to the effect of having a criminal record on employment probabilities among beneficiaries, but existing research suggests that it can significantly reduce chances of employment among the small minority of beneficiaries affected.

Information on the effect of prior criminal convictions on employment among beneficiaries generally was not available in the literature search conducted for this review. It is likely that prevalence rates and the effects of criminal records will vary greatly from country to country (see Western 2001). Based on a survey on very long-term unemployed registered job seekers in New Zealand, Parker (1997) reports that 13% considered prior criminal convictions to be a barrier to employment. The percentage might be higher among discouraged workers who are not registered as unemployed or otherwise not in the labour force.

Criminal convictions and imprisonment can affect employability in a number of ways (see Danziger 1996; Grogger 1995; Western 2001). Any job skills that the individual possessed prior to imprisonment might erode during his or her stay in prison. The psychological and social adaptation a person undergoes while in prison may be incompatible with workplace norms. Having a criminal record, regardless of the individual’s “actual” employability, could send a negative signal to employers about the applicant. Using data from the National Longitudinal Survey of Youth, Western (2001) found that incarceration reduced youth employment by five percentage points. The effect was greater than the effect of dropping out of high school or living in a high unemployment area, and the effect persisted. Fifteen years later, men who had been imprisoned as youths were employed 5–10 percentage points less than their counterparts who did not spend time in prison. The effects of adult incarceration are less persistent, and appear to largely disappear within 4–5 years. However, employment problems after release might draw former inmates into low-wage casual or illegitimate jobs in secondary labour markets, which would then affect their later employability.

2.1.7 Transportation problems

Inaccessible public transportation, lack of car ownership or absence of a driver licence can all act as major barriers for: finding a job, getting to a job every day, getting children or other dependants to arranged care, and dealing with childcare problems or other emergencies while employed. Note that this is a barrier at both an individual level and a community level.

The basic evidence

Numerous studies across countries report that transportation problems are a barrier to employment for all beneficiary types (Danziger and Seefelt 2000; Dawson, Dickens and Finer 2000; Debord et al 2000; Duncan, Kerekere and Malaulau 1996; Fletcher, Garashky and Jensen 2002; Hales et al 2000; Holzer and Stoll 2000; Leete, Bania and Coulton 1999; McKay 2002; McKay et al 1999; Parker 1997; Regenstein, Meyer and Hicks 1998; Social Exclusion Unit 2003; Thomas et al 1999). Mobility problems take the form of a lack of access to public transportation (a community-level issue), lack of access to a reliable vehicle, and lack of a driver licence. Relevant dimensions include access or availability, cost and distance from labour markets (the geography of employment).

Prevalence

Estimates vary according to country and type of beneficiary from a significant minority to a sizeable majority. In other words, mobility barriers are common. In a UK study, 56% of lone-parent Income Support recipients did not have a driver licence, and 28% of those who did had no car available (Hales et al 2000). Two-fifths of UK job seekers report that lack of transportation is a barrier to becoming employed, with one-fourth saying that the cost of transportation is a barrier to attending job interviews (Social Exclusion Unit 2003). In New Zealand, 77% of a sample of very long-term unemployed reported that limited mobility of some form (including lack of access to a vehicle or public transportation and/or not having a driver licence) was a barrier to searching for and taking up work – the highest percentage among all the barriers mentioned – and the percentage was higher among Mäori than among other ethnic groups (Parker 1997).

Nature and significance of the barrier

Research findings suggest that mobility problems inhibit job search and employment sustainability by limiting access to job training centres, places of employment and childcare facilities. Problems reflect individual-level factors (such as low income) that prohibit vehicle purchase, as well as community-level factors (such as a lack of affordable and accessible public transportation). In addition, the nature and significance of transport barriers for beneficiaries will depend on the geography of employment, eg the distance between low-income housing markets and labour markets.

In the UK, research has identified several avenues through which transportation barriers affect the employability of beneficiaries (Social Exclusion Unit 2003). The cost of transport inhibits getting to job interviews and to jobs themselves. Transportation problems keep some young people from even applying for jobs. In a study of low-income families in Iowa, Fletcher, Garashky and Jensen (2002) found that transportation barriers depressed the likelihood of working, controlling for other factors typically associated with employment. Welfare programme managers studied by Danziger and Seefeldt (2000) identified transportation problems (along with educational deficits) as significant barriers to their clients’ employment, but noted that they were not barriers addressed in most “work first” programmes.

Having access to private transportation may facilitate employment to a greater degree than having access to public transportation. Lone parents in the UK working at least 16 hours per week were more likely than those on Income Support to have a driver licence (62% as opposed to 44%) and to have a car available (85% as opposed to 72% of those with a licence), but there were no reported differences between the two groups with regard to access to public transportation (Hales et al 2000). Lone parents in the UK who reported having a driver licence and access to a vehicle in 1999 were more likely than those who did not to be employed at least 16 hours per week in 2000 (McKay 2002). In a Californian study, Cervero, Sandoval and Landis (2002) found that car ownership significantly increased the odds that beneficiaries would make the transition to work but found largely non-significant effects for the quality of public transport.

The geography of employment in a particular labour market plays a large role in shaping the relative importance of public versus private transport. For example, in a study of Cleveland, Ohio beneficiaries, Leete, Bania and Coulton (1999) found that about three-quarters of welfare recipients lived in the central city, while jobs at their skill level were typically located in the suburbs. Thus, a lack of transportation options posed a real barrier. Furthermore, only half of beneficiaries in the central city had automobile access, and those who would have to rely on public transportation faced commute times that were double those of people who could use their own automobiles. Thus, even when access to public transportation is available, the potential costs in terms of time constraints may be inhibitory.

Transportation problems, whether caused by cost, the geography of employment or problems with access or availability, can be both a result of social exclusion and a cause (Social Exclusion Unit 2003). Beneficiaries on low incomes may be unable to afford to buy or maintain a vehicle and may have to live in areas with poor access to public transportation. This then acts as an employment barrier that further reinforces beneficiaries’ social exclusion.

2.2 Family Barriers

2.2.1 Caring responsibilities for children

Age of children and number of children are consistent predictors of being on-benefit or otherwise not working among lone mothers. Cost, availability, lack of knowledge about childcare and the uncertainty associated with starting formal or informal childcare arrangements are all barriers. At least equally important are the attitudes sole parents have about the importance of parental versus non-parental care for their children. The effect is most pronounced when children are young, but parents of teenagers, especially those living in central urban areas, also express concerns about allowing their children to go without parental supervision during work hours.

The basic evidence

Having young children, and having more than two children, may seem like obvious barriers to employment among beneficiaries. Many studies have documented that the presence of children, and the age and number of those children, are key predictors of labour force or educational participation among mothers generally (eg Joshi, Macran and Dex 1996), among lone mothers (Bradshaw and Millar 1991; Colledge 1991; Duncan, Kerekere and Malaulau 1996; Ford, Marsh and Finlayson 1998; Finlayson et al 2000; Holtermann et al 1999), and among welfare recipients (Born et al 1998; Brooks and Buckner 1996; DoL and MSD 2001; Finlayson et al 2000; Kling, Liebman and Katz 2001; Meyers 1993; Noble, Smith and Cheung 1998; Shaw et al 1996). However, the nature of childcare as a barrier is complex. Key dimensions include preferred type, availability, costs and awareness.

Prevalence

Because childcare responsibilities can affect employment in many different and complex ways, identifying overall prevalence rates is difficult. Any information available on the prevalence of the different problems encountered by beneficiaries related to childcare is incorporated below in the discussion on the nature and significance of the barrier.

Nature and significance of the barrier

Responsibility for childcare has been identified as a barrier to employment almost exclusively for the lone-parent beneficiary population. In a New Zealand survey of very long-term unemployed registered job seekers, Parker (1997) found that 17% reported childcare difficulties to be a barrier to employment, with the figure reaching 30% among job seekers from the Pacific population. Just as disabilities among lone parents and the unemployed may be overlooked, it is possible that so too may childcare responsibilities among the unemployed and disabled. This is an area for further research.

Preference for parental care

Concerns about children’s emotional, social and educational wellbeing are primary for many lone mothers, and employment and formal childcare are sometimes seen as compromising children’s wellbeing (DoL and MSD 2001). Thus, the greatest barrier to employment among certain lone mothers may be a strong preference for parental care, a preference they largely share with couple families (Bryson, Ford and White 1997; Marsh et al 2001). As several studies from various countries have found, significant proportions from both groups prefer parental care and, secondarily, care by friends or relatives (Cobb-Clark, Liu and Mitchell 1999; Finlayson et al 2000; Levine, Wyn and Asiasiga 1993; Marsh et al 2001; Millar and Ridge 2001; Schumacher and Greenberg 1999; Stephenson 2001). Cobb-Clark, Liu and Mitchell (1999) found that couple families with a strong preference for parental care achieve this either by having one parent stay home or by altering employment schedules to casual or part-time work (Vandenheuvel 1996). A lone parent must achieve this balance alone.

As Millar (1994) points out for the UK, lone, disadvantaged mothers may have even more reason to want to care personally for their children than other mothers because of their more precarious circumstances. In a US study of low-income families, Kling, Liebman and Katz (2001) found that parents spent a great deal of time protecting their children from the dangers associated with living in a high-poverty, inner-city neighbourhood. Mothers in the sample organised their lives around a need to protect and supervise their children, and this vigilance took precedence over participating in English or high school diploma classes, job training or job search. A lack of trust prevented many from relying on neighbours for supervision.

Although high-crime neighbourhoods might be less common in New Zealand, other aspects of disadvantaged lone parents’ lives might create special needs for parental childcare. As will be discussed in more detail in Section 2.2.2, several US studies show that lone mothers are more likely than other mothers to have a child with a disability or health problem (Olson and Pavetti 1996), which greatly affects their ability to participate in the labour force or in educational/training activities (Brandon and Hogan 2002; Meyers, Brady and Seto 2002). In addition, children experiencing poverty may have greater health problems, which would require more care and increase the need for flexible childcare arrangements that can handle sick children. A study by Abt Associates (1999) in the US indicates that children of mothers on welfare have higher rates of absenteeism from school than other children, including other children experiencing poverty, and a study of these absences in Delaware indicates that most (four out of five) are related to illness.

Together, these examples highlight the need to consider the requirements of older school-aged children as well as younger ones. A New Zealand study found that some lone mothers felt their teenage children required adult supervision to counteract negative peer influences, and thus cited childcare as a crucial factor in their work-related decisions (DoL and MSD 2001).

A distinction should perhaps be made between preferences and attitudes. While preferences for parental childcare may be common, strong attitudes against mothers’ employment may be much more rare. In a UK study of lone mothers by Bryson, White and Ford (1997), only 8% indicated no inclination to be in paid work, and only 11% said that they completely agreed that a mother’s place was in the home. Those with strong attitudes favouring staying at home were much less likely to be employed (15% versus 74% for those with the highest work orientation).

Secondary preference for care by relatives

Lone mothers also express a preference for childcare by relatives because of issues of trust and flexibility. In a study of low-income families in three US cities, Coley, Chase-Landsdale and Li-Grining (2001) report that, compared with mothers using formal care, mothers who used informal home-based care “have less difficulty with transportation or with finding care during their work hours, have care providers who show more understanding of their situation and with whom they share information about their child more regularly, and report that their childcare is more reliable” (p. 4). In a UK study by Stephenson (2001), access to free or cheap informal childcare from friends or relatives was a main differentiating factor between lone mothers who moved from benefit to work and those who did not. This was also the case in a US study (Edin and Lein 1997). However, childcare that meets the needs of the mother in terms of flexibility, communication with provider, and accessibility may not be best for the developmental needs of the child. The type of care that (by far) least meets the developmental needs of the child (unregulated home care) best meets the needs of mothers (Coley, Chase-Landsdale and Li-Grining 2001).

Availability

Beneficiaries, especially the low-skilled, may face additional barriers to securing the type of childcare they are after. High-quality childcare and/or informal care by relatives or friends may not be available during non-standard hours when some casual or part-time jobs take place (Coley, Chase-Landsdale and Li-Grining 2001; Levine, Wyn and Asiasiga 1993). Beneficiaries may live in communities with few resources for high-quality care (Coley, Chase-Landsdale and Li-Grining 2001; Colledge 1991) or in which childcare centres are not within walking distance for those without a car (Levine, Wyn and Asiasiga 1993).

Cost

In addition to preferences for parental care, and a lack of availability of preferred childcare types, the cost of formal childcare may inhibit employment. Costs rise with the number of children and decrease as children get older. Results of studies on childcare subsidies are somewhat mixed. At the macro level, Blank (2000) suggests that the expansion of childcare subsidies, along with evidence suggesting that mothers’ labour force participation responds to the costs of childcare (Anderson and Levine 2000), must not be overlooked as possible factors in the decline in welfare caseloads. Kimmel’s (1995) analysis suggests that lone mothers, particularly white lone mothers, exhibit substantial responsiveness to childcare subsidies, indicating that childcare costs may be a barrier to employment among welfare recipients. Blau and Tekin (2001) also find that childcare subsidies encourage employment and school participation among welfare recipients, but that few receive them, most likely because of insufficient funding and availability.

Awareness

A lack of information on childcare availability and subsidies may also act as a barrier to employment. In the UK, Ford (1996) found that many lone mothers had little information about the availability of formal childcare and its costs and nature. Many lone mothers in their study would have liked to trial a childcare arrangement before beginning employment, but this was not possible. Schumacher and Greenberg (1999), in a review of several US state studies on the childcare arrangements of welfare leavers, found low levels of knowledge about childcare subsidies among recipients moving toward employment. However, even if information is fully disseminated and absorbed, subsidies for formal care may not act as a bridge to employment among sole mothers particularly concerned about parental or familial care for their children. Instead, it may simply induce the already-employed using informal care to switch to formal care in order to take advantage of subsidies (Blau and Tekin 2001).

In summary, although childcare issues are paramount for lone parents, the extent to which they act as barriers to employment will depend in large part on which aspects of childcare are problematic for the beneficiaries involved. For example, those with a strong preference for parental care may face the greatest “barrier” in that their views might mean that they do not consider themselves ready for work, whatever the financial advantages (Millar and Ridge 2001; Ford 1996). For others, the decision to move from a benefit to employment might be much more complex:

They balance the advantages and disadvantages of remaining at home with their children separately from a second equation that balances the gains and losses from work. They balance hours against convenience, quality against cost, avoiding hardship against the quality of parenthood, and many other factors, to arrive at the decision to move towards or away from work.

(Marsh 1997, p. 129)

Research by Finlayson et al (2000) suggests that, once looking for employment, sole mothers will attempt to fit employment around childcare needs, rather than vice versa. For example, one-third of non-employed lone mothers said they would work only if their work hours coincided with school hours. This is an important point, as it could have a large impact on the types of jobs lone mothers are willing or able to move into (eg casual jobs, perhaps representing underemployment for the more skilled) and the likelihood that the employment will be sustainable. It also suggests, again, that it is not only young children that affect lone mothers’ employment.

Longitudinal research on lone mothers in the UK suggests that arranging childcare may be the last barrier that lone mothers need to overcome before starting employment (Marsh et al 2001). This is not to suggest that childcare is not important, but that other barriers may need to be overcome before the practical issue arises of figuring out what to do with the children when the mother is at her job. For example, Marsh et al (2001) found that childcare was rarely the only barrier to employment faced by lone parents not in employment. A large minority faced poor health, disability and low qualifications as well.

2.2.2 Caring for ill, elderly or disabled family members^³

Caring responsibilities are a major barrier to employment for spouses of the disabled and long-term unemployed. Mature-age unemployed workers may have responsibilities for elderly relatives. In addition, children’s health problems and disabilities lower employment probabilities as much as parents’ own health problems and disabilities.

The basic evidence

Caring responsibilities inhibit employment among spouses of the unemployed and disabled (Marsh et al 2001; McKay et al 1999), among the long-term unemployed (Parker 1997; Patterson 1999), and among lone parents who have an ill or disabled child (Baker and Tippin 2003; Brandon and Hogan 2002; Meyers, Brady, and Seto 2002) or who are caring for another adult (Stephenson 2001).

Prevalence

In the US, Meyers, Brady and Seto (2002) estimate that 10–12% of current or recent welfare recipients have a child with a disability. Other researchers have found that, in the US, a higher proportion of families on welfare have a disabled adult living with them than have a disabled or functionally limited child. This affects almost one-quarter of welfare families (Loprest and Acs 1995; Adler 1993). In a New Zealand study by Parker (1997), 19% of registered unemployed job seekers reported being constrained in finding employment by caring responsibilities (13% report childcare responsibilities, but some proportion of those might look after both children and others).

Nature and significance of the barrier

As mentioned previously, given the high rates of poor health and disability among the unemployed and people with disabilities, caring responsibilities for the spouses or other family members of the incapacitated can act as barriers to employment (Marsh et al 2001; McKay et al 1999). In a small UK study of lone mothers and low-income couples, Stephenson (2001) found that caring for an adult was most often a major barrier for the small number of respondents who cited it at all, because of a lack of time or an inability to commit to regular hours. As Patterson (1999) points out, many mature-age workers, who are at greater risk of long-term unemployment, may be caring for elderly relatives. Indeed, in their review of “good practice” aimed at combating age barriers to employment across European countries, Walker and Taylor (1998) include employer flexibility for dealing with elderly relatives in their employment policy suggestions.

Children with ill health or disabilities also present employment challenges for their caretakers. In Stephenson’s (2001) qualitative study, beneficiaries who had a child with a disability were less constrained than those who were caring for an adult because the children’s disabilities did not require as much extra care. However, two larger UK studies identified ill-health of children as a significant factor in preventing work among non-employed lone mothers (Hales et al 2000; Marsh et al 2001). In the US, two-thirds of children with disabilities receiving Social Security Insurance payments had mental impairments, and about two-thirds of these were mentally retarded (Institute for Research on Poverty 1997). It is possible that the type of jobs many low-skilled beneficiaries can obtain are particularly incompatible with caring responsibilities. For example, one US study estimates that only 45% of single mothers who have a child with a disability have access to paid sick leave from their job (Heymann and Earle 1997).

In addition, a US study found that having a disabled child had a substantial effect on mothers’ exits from welfare (Meyers, Brady and Seto 2002). Controlling for a host of factors known to affect welfare exits, the authors found that having a disabled child reduced the likelihood of leaving welfare by as much as did a four-year reduction in the educational level of the recipient. The findings of Brandon and Hogan (2002) corroborate this, putting in on par with the effect of a mother’s own disability. They found that women with children with disabilities were overrepresented among welfare recipients, particularly among the long-term. Importantly, Meyers, Brady and Seto (2002) found that the additional costs of caring for a disabled child were enough to push 4–12% of already poor families into extreme poverty, characterised by food shortages, hunger, evictions and utility shutdowns.

2.2.3 Domestic violence and abuse

Evidence from the US suggests that domestic violence is a potential barrier to employment among lone-mother beneficiaries, but estimates of its prevalence among the benefit population vary widely. When it does exist, its effects are both direct (eg partners sabotaging the efforts of beneficiaries to gain independence by working) and indirect (eg partners lowering self-esteem, morale and confidence). There has been less research on this topic in the other countries covered.

The basic evidence

Most of the studies covered in this review that analysed domestic violence as a barrier to employment were on US lone-mother beneficiaries. The evidence suggests that prevalence rates may be high (Brooks and Buckner 1996; Colten, Cosenza and Allard 1996; Curcio 1996; US General Accounting Office 2001), that experience with current domestic violence may or may not inhibit employment (Barusch 1999; Bassuck et al 1996; Colten et al 1996; Raphael 1996; Lloyd and Taluc 1999; Tolman, Danziger and Rosen 2002), and that experience with abuse in the past or having a history of experience with severe domestic violence is a particularly difficult barrier to overcome (Browne et al 1999; Levine, Wyn and Asiasiga 1993; Tolman, Danziger and Rosen 2002).

Prevalence

Although national-level prevalence of domestic violence among beneficiaries is difficult to estimate, several smaller studies have indicated high prevalence rates among welfare recipients in the US (Brooks and Buckner 1996; Colten, Cosenza and Allard 1996; Curcio 1996). In a review of studies, one US government report suggested that as many as one-third of welfare recipients are current victims of domestic violence (US General Accounting Office 1999). In a sample of 436 homeless and low-income housed women in the US, most of whom were welfare recipients, Brooks and Buckner (1996) found “alarming” rates of violence, with 85% reporting having experienced physical abuse as a child. In New Zealand, Levine, Wyn and Asiasiga (1993) found that some long-term beneficiaries mentioned experience with domestic violence with an ex-partner.

Nature and significance of the barrier

Both current and past experiences with violence can affect employment prospects of beneficiaries but in different ways. Research findings suggest that current abusive partners may sabotage women’s or welfare recipients’ efforts to make the transition to employment, education and training (Bassuck et al 1996; Colten, Cosenza and Allard 1996). Also, current injuries or court appearances can create short-term crises that affect the woman’s ability to leave the home, face prospective employers or take the time to participate in job search (Brush 1999). Such disruptions and injuries can also affect performance on job-training tests, or on the job itself. Over time, an accumulation of injuries and life disruptions – regardless of whether a woman is currently experiencing violence – can affect educational attainment, and create a “cluster of cognitive and emotional symptoms that can impede training and work performance” (Brush 1999, p. 50).

The evidence on current domestic violence as a barrier to employment is mixed. Some studies have documented abusive partners’ efforts to sabotage employment and training activities (Bassuck et al 1996; Colten, Cosenza and Allard 1996). Raphael (1996) found that drop-outs from a Chicago employment training programme were more likely to be experiencing domestic violence than those who stayed in the programme. Other research, using cross-sectional survey data, has not demonstrated an association between current domestic violence and employment (Barusch 1999; Lloyd and Taluc 1999; Tolman, Danziger and Rosen 2002).

Tolman, Danziger and Rosen (2002) point out, however, that cross-sectional data might not provide the right opportunity for specifying the relationship. For example, Browne et al (1999) used longitudinal data on 285 low-income women, mostly welfare recipients, and demonstrated that past experience with domestic violence affects subsequent job stability. Women who reported experiencing physical aggression were one-third as likely to be employed at least 30 hours per week for six months or more in the subsequent 12-month period as women who did not report such violence. In New Zealand, Levine, Wyn and Asiasiga (1993) found that past experience with domestic violence among long-term beneficiaries, combined with long-term benefit receipt, lowered women’s confidence levels about finding employment or taking up training opportunities. Building on the work of Browne et al (1999), Tolman, Danziger and Rosen (2002) used two waves of survey data from the Women and Employment Survey in Michigan to study trajectories of violence and employment. They found that violence that is severe, recent and persistent (long term) is associated with welfare reliance and a lack of employment. Women experiencing this type of violence (characterised by all three dimensions) were about twice as likely to be welfare reliant as women who had never experienced domestic violence. Thus, it may be that both past and current experiences with violence combine to create particularly high barriers.

2.3 Social and Community Barriers

2.3.1 Lack of access to or poorly developed social networks

One of the most successful job search strategies involves the connections of families and relatives, and a person lacking in the right social networks can be disadvantaged in the labour market. In addition, social networks can provide access to employment support, such as childcare and transportation, as well as socio-emotional support.

The nature and extent of beneficiaries’ social networks can have an important effect on their ability to find and keep employment. One Canadian study found that “social capital” in the form of social networks was the most significant factor in predicting welfare exit among a sample of long-term beneficiaries (Levesque and White 2001).

A number of factors influence the development of social networks, including an individual’s employment history (which, over time, contributes to the number of people the individual has had contact with), the formation of work-based friendships and the individual’s availability for civic participation. Geography is also important. The size and type of neighbourhood in which a person lives determines in part the potential for social contact, as well as the nature of that contact – a community with high proportions of jobless individuals and families will not provide opportunities for connections with friends and acquaintances with links to the labour market (Wilson 1987). Geography matters for other reasons as well, most notably in terms of local economic conditions and employment opportunities (see Section 2.4.2). However, in this section, the focus is on social conditions, mostly at the community level, that can help or hinder a beneficiary’s chances of being employed.

Nature and significance of the barrier

Social networks and instrumental aid

Friends and relatives can provide leads on employment opportunities and assistance with employment-related needs, such as childcare and transportation. Indeed, research indicates that social networks can be vital for successful job search. Several studies in the UK have found that at least one-third of jobs are found through such personal contacts (Bottomley, McKay and Walker 1997; McKay, Walker and Youngs 1997). One US study found that social networks are particularly critical in the hiring process of ethnic minorities, but that many job seekers lack access to the social networks that lead to employment or use them less well (Petersen, Saporta and Seidel 2000).

Social networks shrink once an individual leaves employment (Kempson 1996; Williams and Windebank 2000). Gallie (1994) found large discrepancies between the social networks of the employed and the unemployed. Long-term unemployed people tend to associate mainly with others who are long-term unemployed (Kempson 1996), which diminishes potential “word-of-mouth” connections to possible employment. Using British Household Panel Study data, Hannan (1998) found that unemployed individuals with at least some close friends who were employed had significantly higher employment probabilities than those with no close friends who were employed. The effect remained strong, even after controlling for a host of other variables known to affect employment probabilities, like employment history and local employment conditions.

Social networks can provide other types of instrumental aid that facilitate employment. In their US-based study of former lone-mother beneficiaries, Edin and Lein (1997) found that the most successful wage-reliant women were those whose “hidden” costs of employment were offset by a supportive social network that provided free or inexpensive childcare, monetary support or access to transportation. Given parents’ preferences for informal care by friends and relatives for their children (see Section 2.2.1), access to social networks that provide this may be especially important for facilitating employment among lone mothers, who do not have a partner to rely on. In Hannan’s (1998) UK study, 10.8% of the unemployed, compared with only 3.8% of the employed, perceived that they had no one to rely upon in a crisis. This figure increased to 20% among those unemployed in Year 1 and still unemployed in Year 5. This is in contrast to a figure of 9.5% of those who had since become employed (Hannan 1998).

Social networks and social support

In addition to instrumental aid in finding and keeping employment, social networks can provide critical social support to beneficiaries, although this is an area requiring further research (Hannan 1998; Sansone 1998). The general psychological literature suggests that being connected with supportive others can enhance self-esteem and feelings of self-mastery and control (Antonucci and Akiyama 1987), all of which are factors that support finding and keeping employment. Hannan (1998) found that the unemployed have higher levels of psychological stress than the employed. Importantly, the unemployed whose close friends are all employed have much lower psychological stress levels than those who have only some or no close friends who are employed. This finding suggests that the effect of having a “good” social network in terms of employment may be the result of positive social support rather than just instrumental aid. Inadequate social support is associated with greater risk of depression among mothers on welfare in the US (Coiro 2001) and with an inability to cope with low income and feelings of social exclusion among low-income mothers in Scotland (Scott et al 1999). As is discussed in Section 2.5.2, the social support provided by effective case managers has been identified as a critical component in several successful employment and training schemes for beneficiaries in both the UK and the US (DeBord et al 2000; Evans 2000; Lewis et al 2000; Sansone 1998; Woodfield and Finch 1999).

In summary, social networks can be important sources of information about and support for employment. Indeed, community-based employment schemes that recognise the importance of social networks in facilitating employment opportunities among beneficiaries have been successful at lowering welfare dependency and social exclusion, particularly among the long-term unemployed (Evans 2000; Foreman, Hawtin and Ward 1997).

2.3.2 Geographical location

As noted in Section 2.3.1, the place where a beneficiary lives has an impact on access to childcare, public transportation, social networks, social support and employment opportunities. All of these issues are discussed individually in the various sections in this review, but it is worth noting that this one factor can have an impact on many of the barriers discussed in this report. Beneficiaries in rural communities may face a distinct set of barriers relating to their geographical location. Several New Zealand studies, in particular, mention rural – and in some cases suburban – location as a barrier to accessing employment and educational opportunities (Duncan, Kerekere and Malaulau 1996; Parker 1997; Stewart et al 1998). Ethnic differences in location can then lead to ethnic differences in barriers to employment (Parker 1997). In addition, anecdotal evidence suggests that a lack of resources might lead beneficiaries to move to areas with low costs of living, which then put them further out of reach of employment and educational opportunities.

Concentrated disadvantage and “neighbourhood effects”

A growing concern in the countries in this review is the increasingly geographical concentration of disadvantage (Callister 1998; Glennerster et al 1999; Jargowsky 1997; Morrison 1993; Soldera 1999; W.J. Wilson 1987, 1996). This was not an issue analysed in any depth in the studies covered by this review, but it is one that it is important to note for policy-making purposes.

Studies in New Zealand (Callister 1998; Soldera 1999) and Australia (Hunter 1996) have indicated that local areas are becoming more socially stratified with, for example, gains in education going disproportionately to areas with already-high concentrations of well-educated individuals. In Wellington, Morrison (1993) found that Mäori with higher levels of education are likely to live in or move to areas with educationally similar residents. However, as Hunter (1996) shows for Australian cities, changes or differences in the personal characteristics of individuals within local areas do not alone explain area-level differences (or changes over time) in employment. Instead, he concludes, “geography matters”.

Wilson’s (1987) thesis on “concentration effects” posited that the effects of being poor and living in a poor neighbourhood go beyond their simple additive effects. Instead, as Massey (2001) clarifies, their effects are interactive and compounding. The reasons for this effect have been the subject of great debate, with some scholars pointing to a “culture of poverty” that results in internalised norms that are not conducive to employment (Mead 1986), while others emphasise the effects of social isolation from mainstream institutions (Wilson 1987). More recent work has been aimed at specifying the mechanisms through which neighbourhood effects operate to affect individuals and families (see summary in Avenilla and Singley 2001, and individual chapters in Booth and Crouter 2001). In any case, the general finding that local areas matter, above and beyond the effects of individuals’ characteristics, implies that policies need to be aimed not only at individuals but also at communities (Hunter 1996).

2.4 Work-Related Barriers: Individual and Structural

2.4.1 Human capital issues

The basic evidence

Beneficiaries of all types face distinct disadvantages in the labour market because of a lack of job skills (both basic and job-specific) (Danziger and Seefeldt 2000; Danziger et al 2000; Parker 1997; Riccio and Freedman 1995), a lack of work experience (Danziger et al 2000; Danziger and Seefeldt 2000; Loumidis et al 2001; Parker 1997; Riccio and Freedman 1995; Vartanian and McNamara 2000), and low education levels (Danziger and Seefeldt 2000; Zedlewski and Aldersen 2001; Loumidis et al 2001; Marsh et al 2001; Vartanian and McNamara 2000; Parker 1997). All three types of barriers – job skills, work experience and education level – are intertwined. Those with less work experience are less likely to have either basic job skills or job-specific skills. In addition, the relative importance of the three types of barriers depends heavily on the nature of labour demand (see also Section 2.4.2). In an economy with a high demand for high-skilled labour, a lack of work experience will be less important than a lack of education. However, having low levels of education, job-specific skills and work experience will matter less than having basic job skills in a local economy that is heavily dependent on unskilled labour. Generally speaking, poorly educated workers face barriers to employment in all of the (increasingly high-skill) economies covered in this report (see Section 2.4.2).

Prevalence

In the US, 44% of lone-parent welfare recipients in 1999 lacked a high school degree or its equivalent (Zedlewski and Aldersen 2001). In Australia, people with disabilities have lower educational levels than the general population, and the evidence suggests that a childhood disability lowers educational attainment (Bradbury, Norris, and Abello 2001). In the UK, one-third of non-working lone parents had no vocational or academic qualification (Marsh et al 2001). Among very long-term unemployed job seekers in New Zealand (Parker 1997), 11% reported that a lack of qualifications was the major barrier to employment. This was the third most common response, following limited mobility and age discrimination.

In the New Zealand study, the fourth most common barrier cited by long-term unemployed job seekers was a lack of appropriate work skills or work experience. Indeed, for long-term beneficiaries, a lack of recent work experience is, by definition, a key barrier. In a recent UK study of low-income families, only 15% of lone mothers and 34% of couples had worked 16 or more hours per week at some stage during the last two years (Marsh et al 2001).

A lack of both basic and more specific job skills is common among beneficiaries, particularly the long term. Pavetti (1999) found that among long-term welfare recipients, almost two-thirds scored in the lowest decile on a cognitive-ability test. A review of several US studies suggests that 30–45% of all welfare recipients lack a high school degree, and 20–30% lack basic job skills (US General Accounting Office 2001).

Nature and significance of the barrier

At the individual level, beneficiaries with better qualifications are more likely to find employment, and to find it more quickly and then keep it (see Athanasou, Pithers and Peoumenos 1995 on long-term unemployed in Australia; Brooks and Buckner 1996 on lone mothers in US; Jonczyk and Smith 1990 on disabled in Australia; Levine, Wyn and Asiasiga 1993 on lone mothers in New Zealand; McKay, Walker and Youngs 1997 on unemployed in UK). For example, McKay, Walker and Youngs (1997) found that the possession of qualifications more than halved the median duration of unemployment, and the effect was stronger for men than for women.

In addition, research has shown that education and training programmes aimed at eventually moving lone-parent beneficiaries into employment have positive effects on their employment rates in the US (Riccio, Friedlander and Freedman 1994; US General Accounting Office 1999) and the UK (Ford, Marsh and Finlayson 1998). Increased educational levels of single mothers (along with improved area conditions) accounted for the major share of gains in their wellbeing between 1993 and 1999 in the US (Mills, Alwang and Hazarika 2001).

A lack of basic or relevant job skills may also be a barrier. Long-term beneficiaries are particularly likely to have low skill levels. In the UK, adults with low basic skills levels (particularly numeracy) are five times more likely to be unemployed than those with average skills (Sparkes 1999). As Bryson, Ford and White (1997) point out, a lack of time and energy to develop skills while being a lone parent may be the reason that the duration of lone parenthood is negatively associated with hourly earnings. In addition, becoming a lone mother at an early age can have substantial negative impacts on human capital accumulation across the adult life course (Klepinger, Lundberg and Plotnick 1997). Low education and skill levels can also affect participation in welfare-to-work programmes. Pavetti et al 1997 report that, in Canada’s Self-sufficiency Project, most of the least educated lone mothers who were eligible for an earnings supplement programme did not sign up to receive the supplement.

A lack of work experience can affect both time on welfare and employment retention for those who do make a transition to employment. In a US study of a local labour market, having a steady employment pattern up until the beginning of benefit receipt was associated with spending less time on welfare (Coulton et al 1997). In a study of former welfare recipients employed at firms in four different US metropolitan areas, those with less work experience had greater difficulty retaining employment (Holzer, Stoll and Wissoker 2001).

Corcoran and Loeb (1999) show that even low-skilled workers and long-term welfare recipients receive substantial wage returns to work experience. However, wage growth is slower for part-time work. Thus, the authors conclude that policy efforts aimed at encouraging full-time work amongst beneficiaries will pay off in the long run for former welfare recipients in terms of their ability to earn a living wage.

While individual-level deficits in education, skills and experience are important barriers, it is also necessary to consider how these human capital attributes interact with broader structural factors such as the nature of labour demand in the macroeconomy and the local economy.

2.4.2 Labour demand issues

Studies of barriers to employment among beneficiaries often focus rather narrowly on individual- and family-level problems rather than on structural and contextual factors that also restrict opportunities for employment. One of the most important of these structural factors is the demand for their labour, which has both national and local dimensions (see Glennerster et al 2000 for discussion of their relative importance). Although a thorough review of labour demand issues is beyond the scope of this review, a few key points are mentioned here in order to provide a context for assessing the importance of the other barriers discussed in this report.

Macroeconomic demand for labour

Both the overall demand for labour (as reflected in the unemployment rate) and the type of skills demanded by the economy as a whole are important when considering the employment prospects of beneficiaries. As many researchers have noted, the demand for unskilled and low-skilled labour declined in the period immediately preceding the most recent benefit system reforms in the countries covered here, causing some concern about the outcomes associated with welfare-to-work programmes (Currie et al 2001; Webster 2000; Wilson 1998).

Studies in several countries suggest that the restructuring of the economy toward high-skill jobs on the one hand and casual, temporary and part-time jobs on the other has particularly disadvantaged beneficiaries with disabilities. In a study on incapacity beneficiaries in the UK, Berthoud (1998) suggests that the rise in incapacity benefit claims can be tied to changes in the labour market, in which employers have become more selective in their hiring practices. Research by Skogman (1998) suggests that a similar process has occurred in Sweden, where decentralisation of wage-setting practices during the 1980s allowed employers to base wages on their assessment of individual ability, which they consistently underestimate for applicants with disabilities. Barnes, Thornton and Campbell (1998) identify several changes in the UK labour market that have further disadvantaged the employment prospects of incapacity beneficiaries, and many of their ideas apply to other beneficiaries as well.

The current labour market puts a high premium on education, which is difficult for people with disabilities to access and tends to be lower amongst the beneficiary population in general. Also, employment of people with disabilities has historically been concentrated in unskilled jobs, which are shrinking, while growing service-sector and team-based working environments can place a premium on communication skills, which may be lacking in those with speech or hearing problems. In New Zealand, Lunt and Pernice (1999) also suggest that people with disabilities will be particularly disadvantaged in the restructured economy, which has increased casual, temporary, part-time and service-sector employment.

The types of jobs being created (eg part time, temporary, casual) may conflict with the needs of many beneficiaries. In a study of the unemployed in the Auckland eastern suburbs, Stewart et al (1998) found that only 25% would consider taking a part-time job, suggesting that the increase in the demand for part-time labour in New Zealand in the 1990s may have created a further mismatch between labour demand and supply. Sole mothers may prefer flexible part-time employment, but they also place a high value on job (and income) security (DoL and MSD 2001), which is often lacking in the “flexible” jobs that are increasing in the low-skilled end of the labour market.

In the US, concerns about the negative impact of changes in skill-specific labour demand were countered by subsequent improvements in the economy overall that increased the demand for labour generally. Indeed, the unprecedented sustained economic growth during the 1990s in the US has been credited with making it possible for so many poorly skilled and poorly educated beneficiaries to find employment during the era of welfare reform (Blank 2000; Council of Economic Advisers 1999; Holzer 2000; Meyer and Rosenbaum 2000; Moffitt and Stevens 2001). Low unemployment levels benefit the poorly skilled because employers are forced to rely on less traditional sources of labour and to provide training and opportunities for the less skilled (Blank 2000). Summarising available evidence, Blank (2000) states that between one-third and one-half of the decline in welfare caseloads since 1996 can be attributed to macroeconomic factors, and suggests that these factors have worked in tandem with changes in policy to increase employment among welfare recipients. Available evidence also suggests that the economy played a large role in welfare caseload declines during the early 1990s (Moffitt and Stevens 2001). In addition, after wages of the less skilled declined between the 1970s and early 1990s, there was significant wage growth among this group during the late 1990s, further improving employment prospects for former beneficiaries (Blank 2000). However, the more recent downturn in the economy, coinciding with the end of welfare time limits for the first cohort affected by them, has renewed concerns about the employment prospects and wellbeing of former beneficiaries.

Local labour market demand

Changes in the demand for labour have had differential affects on local areas. Increasing demand for highly skilled workers has been coupled with geographical pockets of severe decline in the demand for unskilled and low-skilled employment in, for example, the manufacturing sector (Webster 2000; Kasarda 1996). Some local areas can therefore have severe “job-skill mismatches” that make it difficult for beneficiaries to find employment, and these mismatches go beyond whatever changes are occurring in the nomy. Empirically, Holzer and Danziger (1998) estimate that job availability in the US was three times higher for women in general than for welfare recipients when the jobs available were “matched” to potential employees in terms of skills, location and race. They estimate that 20–40% of welfare recipients would have a difficult time finding a job in the local labour market, even in the absence of accounting for physical and mental disabilities or substance-abuse problems.

More generally, local labour market conditions have been identified as important factors in movements into employment among the unemployed (Iacovou and Berthoud 2000; McKay, Walker and Youngs 1997), although Millar and Ridge’s (2001) review of the UK evidence suggests that research findings are mixed. This may be because of inadequate measurement. Overall, it appears that the academic research on labour demand more generally has not been well integrated into policy-related studies on barriers to employment among beneficiaries.

Barriers arising from the structure and functioning of the labour market

Individuals in certain subgroups of the population may face additional barriers to employment because of the structure and functioning of the labour market. As M Wilson (1996) describes, different theories of labour market functioning all predict that the young, the low skilled, ethnic minorities and older workers will be vulnerable to unemployment but for different reasons. For example, the standard neoclassical view is that all of these groups are more likely to be unemployed because of their lower levels of human capital (as discussed in Section 2.4.1) or because of a preference for leisure (eg among workers closer to retirement). Institutional theories, which draw on insights from the sociological literature, highlight the importance of structural features of the labour market that differentially reward certain population subgroups depending on characteristics that go beyond human capital (Farkas and England 1994). These theories suggest the importance of issues such as discrimination (see Section 2.4.3), social networks (see Section 2.3.1) and job queues (M Wilson 1996).

As a number of studies have demonstrated, the demand for labour can vary according to individuals’ ethnicity, gender or immigrant status because of the historical structuring of the labour market (Brosnan and Wilson 1989; Reskin and Roos 1993; Waldinger 2003). Thus, for example, female beneficiaries face the same labour market disadvantages that all women face, and this includes high levels of occupational segregation and issues of comparable worth (Colledge 1991; McHugh and Millar 1996; Singley 1995). This affects not only their employability once they become lone mothers, but also their work histories prior to becoming lone mothers. McHugh and Millar (1996) describe an evaluation of an employment and education programme for lone mothers in Australia that found that most clients had work histories concentrated in female-dominated jobs, characterised by limited skills, low pay, few career prospects and poor job security. Similarly, clients who left the programme for work typically found jobs in female-dominated occupations, with a substantial minority in unskilled and insecure positions (Shaver et al 1994). Thus, social and labour market policies aimed more broadly at structural inequalities (eg equal pay or anti-discrimination legislation) may address some more general barriers for the beneficiary population.

2.4.3 Actual and perceived discrimination

Evidence suggests that both perceived and actual discrimination in terms of beneficiary status or other characteristics may act to inhibit job search and hiring. A New Zealand study of long-term unemployed identified perceived age discrimination as one of the major barriers.

The basic evidence

Discrimination, both perceived (Hirsch 1999; Jonczyk and Smith 1990; Parker 1997; Scottish Poverty Information Unit 1998) and actual (Encel 2000; Graham, Jordan and Lamb 1990; Kennelly 1995; Kirshenman and Neckerman 1991; Lewis et al 2000; Lunt and Pernice 1999; Ravaud, Madiot and Ville 1992), is frequently cited as a barrier to employment, although its importance varies by beneficiary type. The long-term unemployed and those with disabilities may be most affected.

Nature and significance of the barriers

Perceived discrimination

Several studies have implicated perceived age discrimination in depressed job search among the long-term unemployed and people with disabilities (especially those who are male and over the age of 45) (see Gregg, Johnson and Reed 1999 for unemployed in the UK; Hirsch 1999 on disabled in the UK; Jonczyk and Smith 1990 on disabled in Australia; Parker 1997 in New Zealand; Patterson 1999 on unemployed in New Zealand; Scottish Poverty Information Unit 1998 on disabled in Scotland). Age discrimination was a barrier identified by large percentages of both young and older very long-term job seekers in New Zealand, including 85% of those over 50 (Parker 1997). Based on his experience working with the mature-age unemployed in New Zealand, Patterson (1999) also suggested that age discrimination among employers is a major barrier. For males with disabilities who are over 45, their medical condition itself becomes less of a barrier, and age discrimination and employer attitudes toward people with disabilities become more important. In Australia, 90% of disabled clients over 45 identified age discrimination as a major barrier, and a majority reported experiencing age or disability discrimination first-hand. Many had been laid off initially because of their disability (Jonczyk and Smith 1990).

Actual discrimination

The extent to which discriminatory practices actually affect overall employment probabilities for beneficiaries is difficult to assess. Instead, some studies, primarily of the mature-age unemployed, rely on indirect evidence. For example, Campbell (1999) shows that unemployed older people (aged 55–65) have very low chances of ever working again. For both men and women in their late 40s, their chances of working again are about 50%, even under favourable conditions. Although other problems probably contribute to these odds (eg poor health and disability), the author suggests that age discrimination is a real problem. Gregg, Johnson and Reed (1999) find that males over 50 are disproportionately represented among the long-term non-employed in the UK, and Statistics New Zealand (2002) shows that, among the unemployed, those 45 and older have the highest representation among the long-term unemployed.

Other studies focus more directly on employer hiring practices. Employers interviewed in the UK expressed prejudiced views about lone parents as employees, suggesting that they may be less flexible and available than other employees (Lewis et al 2001). Kennelly (1995) found similar employer prejudice against black single mothers in Atlanta. One recent US study of 750 employers found wide demand for welfare recipients as employees, but the results suggested that employer demand for these workers may depend on the tightness of the labour market (Holzer and Stoll 2000). Similarly, an Australian study of mature-age workers found that employers make decisions about hiring based on the typical age of the customers served and the age-associated qualities they are looking for (eg experience versus risk-taking) (Bittman 2001).

Encel (2000) states that the principal barrier to employment among mature-age, long-term unemployed beneficiaries in Australia is age discrimination on the part of employers, who consistently give preference to younger workers. Two UK studies and one US study suggest that employers are willing to overlook any skill deficits of beneficiaries (many of whom are older workers in the UK studies) if they have positive, motivated attitudes (Bunt, Shury and Vivian 2001; Regenstein, Meyer and Hicks 1998; Snape 1998). However, these qualities may be lacking in those who have spent long periods of time without work.

Studies by Graham, Jordan and Lamb (1990) and Ravaud, Madiot and Ville (1992) show that employers prefer to hire a person without a disability over a person with a disability who has the same qualifications and characteristics. Policy interventions have been developed in New Zealand to provide incentives for employers to hire people with disabilities; however, Lunt and Pernice (1999) report anecdotal evidence suggesting that employers do not retain these employees once their subsidies run out, leading to employment instability. A UK study of employers who had hired former beneficiaries found that most were happy with the placements. However, those who did have problems reported poor attendance, particularly among those with physical disabilities, and a lack of confidence or motivation, most often found among the formerly long-term unemployed (Bunt, Shury and Vivian 2001).

A number of studies have documented that employers may prefer hiring workers on the basis of sex or ethnicity. For example, applicants may be scored lower when applying for jobs typically undertaken by members of the opposite sex (eg a woman applying for a job in a male-dominated occupation may score lower than a similarly qualified man) (Davison and Burke 2000). Stewart and Perlow (2001) found no differences in employers’ propensity to hire blacks versus whites overall, but did find that evaluators with more negative attitudes toward blacks were more confident in their decision to give black applicants low-status jobs and white applicants high-status jobs, despite them being equally qualified. In addition, firms that are owned by or serve members of a particular racial or ethnic subgroup may favour hiring employees of similar race or ethnicity (Raphael, Stoll and Holzer 2000).

2.4.4 Ineffective job search

Some evidence exists that beneficiaries may be unaware of the most common job search methods used by local employers, creating a mismatch that inhibits employment.

As discussed in Section 2.3.1, friends and relatives can be key sources of information for job searchers, and beneficiaries without social networks will be less effective at finding a job. In addition, beneficiaries may be unaware of the methods most commonly used by the types of employers most likely to hire them. Thus, a mismatch between employer recruitment practices and the job search methods of the non-employed can create barriers to finding employment. Barnes, Thornton and Campbell (1998) report on a study of employers in North London that showed that, while employers typically used recruitment agencies for hiring, job seekers with disabilities typically used job centres.

Canada’s Self-sufficiency Project (Greenwood and Voyer 2000) provides powerful evidence that assistance with targeting job searches can go a long way in helping beneficiaries into employment. One experimental group received only financial incentives for taking up full-time employment, while another was given the incentives as well as comprehensive assistance with job search (eg help with preparing resumes, a job-finding club, job leads and job coaches). One-third of participants in the group offered only financial incentives took up full-time jobs within a year, compared with over one-half of those who received both the incentives and job search assistance.

2.5 Benefit System Barriers

2.5.1 Financial disincentives to employment

The issue of financial disincentives is complex because of the interaction of wages, taxes and benefits. Aggregate trends in benefit receipt do not clearly follow benefit levels, and several studies show that recipients have very low reservation wages, given their preference for working versus staying on-benefit. However, there is evidence that the transition costs associated with starting employment are inhibitory. Evidence suggests that earnings supplementation (in-work benefits) and other reforms aimed at decreasing marginal tax rates are having an effect on boosting employment levels, even (and, in some cases, especially) among long-term beneficiaries.

Financial disincentives arise when a person or family is financially better off receiving out-of-work benefits and not working than they would be (or than they perceive that they would be) if they started employment. Financial disincentives are not only, or even mostly, a function of the benefit system. Some arise directly out of the benefit system. These are related to benefit levels and the proportion of benefits that are withdrawn with the take-up of employment. Other disincentives arise from the nature of low-wage employment, and are influenced by beneficiaries’ education and skill levels and local area conditions, or the costs of taking up and keeping employment. However, all are interrelated and are therefore discussed together in this section.

The research evidence on financial disincentives comes from three types of sources:

interviews with beneficiaries that probe their decision-making processes regarding work
microsimulation models analysing the effect of various hypothetical changes in benefits, wage rates and taxes
evaluations of the effect of various real-life changes in both out-of-work and in-work benefits and tax credits.

Generally, the evidence is mixed on the importance of financial disincentives as barriers to employment among beneficiaries, but it does suggest that certain types of disincentives (primarily transition costs) are more important than other types.

Benefit levels and employment

One way financial disincentives to employment arise is if benefit levels are high. In the absence of other factors (like a preference for work), there may be little incentive to work for a similar or lower level of income. The relationship between benefit income and after-tax earnings is called the “benefit replacement rate”. The combination of the rate of withdrawal of benefits as a beneficiary enters employment (“benefit abatement”) and taxes on their earnings can create high Effective Marginal Tax Rates (EMTRs) that diminish the attractiveness of working for similar levels of income to that which they received while on-benefit.^⁴If the only employment available is low-wage and/or insecure as a result of local conditions or human capital deficits, employment itself may not offer enough financial rewards for an individual to leave the benefit.

Targeted studies generally find that there are mixed effects of benefit levels on employment. Two studies, one from the US and one from the UK, suggest that there is no clear relationship at the aggregate level between benefit levels and employment levels of “at-risk” groups (lone mothers and former beneficiaries). For example, in the UK during the 1980s, benefit levels and the labour force participation rate of lone mothers both fell (Bryson, Ford and White 1997). US national-level data from 1969–1993 show that former beneficiaries were no less likely to work and no more likely to be back on welfare in states with high welfare payments (Vartanian and McNamara 2000). In a study of lone mothers in 1991, Bryson, Ford and White (1997) estimated that the employment probability of a lone mother receiving 10% more in Income Support was only four percentage points lower, all else being equal. The authors estimated that, by 1995, Income Support levels had no effect at all on the employment probabilities of 60% of lone mothers.

Review articles on the marginal effect of benefit income on labour supply generally state that effects are small or non-existent (see Brosnan, Wilson and Wong 1989). It appears that the most significant effect of even large increases in benefit income would be a drawing out of the job search process rather than a disincentive to engage in employment (Blundell 1994; Brosnan, Wilson and Wong 1989). In contrast, the unusual quasi-experimental nature of unemployment insurance reforms in Sweden allowed Carling, Holmlund and Vejsiu (2001) to determine that benefit reductions led to a 10% increase in employment transition rates, which the authors characterise as large compared with results from previous studies. Also, Maloney (1997) attributed 40–80% of the increase in total employment during the 1990s in New Zealand to the 1991 cuts in benefit rates.

Benefit systems that withdraw support when one partner of an unemployed couple takes up employment can create a perverse effect whereby the couple is better off financially by remaining on-benefit. However, the empirical evidence suggests that the effect is likely to be small. For example, in the UK, McKay, Walker and Youngs (1997) found that, while employment levels among partners of the unemployed were lower than among partners of the employed, the benefit system effects explained only 11% of this difference. In addition, among partners who had been employed when their spouse became unemployed, only 4% changed their employment status because of their partner’s unemployment (McKay et al 1999). Two-thirds of non-employed partners of the unemployed reported family reasons for not working, and only 14% said that they were not in paid work because their partner’s benefit would be reduced (McKay et al 1999).

Low earnings and the costs of making a transition to employment

There is more evidence to suggest that low earnings from jobs, as opposed to high benefit levels, act as a barrier to employment. The effect is exacerbated by benefit systems that quickly withdraw income support benefits or other assistance (eg housing benefits) and by high transition costs (some temporary and some permanent) that can lower a given wage rate’s worth compared with the current benefit level.

A number of studies, particularly from the UK, the US and New Zealand, identify both the perceived risk and the reality of the costs involved in taking up employment as a barrier for some beneficiaries. Costs identified among all types of beneficiaries include: income instability while transitioning from benefit income to earnings income; income shortages due to loss of other benefits such as housing assistance; and income shortages due to new work-related expenses such as transportation, childcare, footwear and union dues (Dawson, Dickens and Finer 2000; Edin and Lein 1997; Finch et al 1999; Hales et al 2000; Levine, Wyn and Asiasiga 1993; Lewis et al 2001; Loumidis et al 2001; McKay, Walker and Youngs 1997; Parker 1997; Puniard and Harrington 1993; Stafford et al 1998; Thomas et al 1999). Concerns about the financial costs of working are all the more relevant for low-skilled or unskilled beneficiaries who may face a labour market consisting of casual, temporary or otherwise unstable jobs. For them, the expenses associated with the transition may never be recovered, and the risks of leaving the stability of public assistance are very high (Bryson, Ford and White 1997).

At the micro-level, Parker (1997) found that 36% of New Zealand’s very long-term unemployed workers who were surveyed said that they would have a hard time earning more than they received on-benefit. This perception was particularly prevalent among those registered as job seekers for more than two years. In a New Zealand evaluation study of benefit reforms, researchers found that beneficiaries who had taken up part-time employment felt that the financial gain was only marginal, given start-up and ongoing costs of work and loss of income as a result of debt or abatements (DoL and MSD 2001).

Studies have found that anticipatory perceptions regarding these costs may be as important as the costs themselves as barriers to sustainable employment (Millar and Ridge 2001). Job seekers who expressed concerns about fluctuating income levels during the transition to work were more likely than those who did not to stay on‑benefit (McKay, Walker and Youngs 1997). Levine, Wyn and Asiasiga’s (1993) study of lone parents in New Zealand found that some were worried about covering their expenses between the last benefit payment and the first pay cheque. Almost half of very long-term unemployed job seekers in New Zealand said that they could not afford to buy the work clothes or other personal goods needed to start work. They also said (Parker 1997), as did incapacity beneficiaries in the UK (Loumidis et al 2001), that they would feel more confident taking up training or employment if they knew that they could return to the benefit if the job did not work out. Similarly, the uncertainty of knowing how childcare would work out once they started work led many lone mothers to express a desire to trial childcare before starting employment (Ford 1996). In terms of income insecurity, it is worth noting that the Working Families Tax Credit (WFTC) in the UK is designed to address this issue by allowing recipients to receive the credit for six months after starting work, even if the job does not work out.

In studies of lone parents (Hales et al 2000) and job seekers (McKay, Walker and Youngs 1997), there were fewer actual transition problems encountered among those who took up employment than were anticipated by those who did not. Even so, almost one in eight of those who took up employment reported experiencing financial difficulties associated with losing housing and other assistance, and one in five reported income shortages (McKay, Walker and Youngs 1997). In a later study of the unemployed, one-fifth of those who took up employment reported financial problems in the transition, most commonly waiting for wages to be paid and meeting increased living costs (McKay et al 1999). In a study of Australian sole parents and unemployment beneficiaries who took up employment, Puniard and Harrington (1993) found that only 6% reported not having any work-related costs. Sole mothers were more likely to report high work-related costs because of childcare expenses and, to a lesser extent, loan repayments for cars that were needed to commute to work. (In contrast, couple families avoided high costs of childcare by having one partner stay home or by arranging their work schedules around childcare needs.) Notably, respondents also felt that new time constraints increased living costs, such as the need to buy lunches or takeaway dinners. Hales et al (2000) studied lone parents in the UK who moved into employment from Income Support and found that many encountered financial difficulties associated with increased living expenses, as did lone mothers in Edin and Lein’s (1997) US study.

Taking into account beneficiaries’ concerns about the transition into employment, Finlayson and Marsh (1998) suggest that lone parents have modest wage expectations but high “reservation job-security”, ie the steadiness of the income stream was a more important concern than low wages per se. In a review of the evidence on financial disincentives in the UK, Gardiner (1997) concludes that “poverty traps” and “unemployment traps” are probably less important as barriers to employment than previously thought. Instead, she concludes that the uncertainty of employment and the insecurity associated with transitioning into employment are much more serious barriers. She posits that beneficiaries with higher living costs (eg lone mothers, people with disabilities or parents of children with disabilities) do not have higher reservation wages but instead have higher needs for income security.

The effect of earnings disregards, earnings supplements and tax credits

A number of reforms have been aimed at increasing incentives to take up low-wage employment. Some, like earnings disregards, operate within the existing benefit system, while others, such as earnings supplements and tax credits, work outside of it (Blank, Card and Robins 1999). Studies that focus on the effects of particular programmes or tax credits aimed at increasing the financial advantages of employment and easing transition costs generally indicate that these reforms are reducing barriers and increasing employment among beneficiaries, although effects vary in important ways.

State-level data in the US suggest that states that adopted more generous earnings disregards prior to 1996 showed slightly higher rates of employment among beneficiaries in 1997 than states that adopted them later or not at all (Blank, Card and Robins 1999). Stafford et al (1998) report that housing assistance during the transition to employment acted as a financial buffer and assisted in moving claimants into employment. Former beneficiaries who used the housing assistance reported that they would not have accepted an employment offer so readily without the assistance. Experimental evidence of a Minnesota programme indicated that earnings disregards alone were not very effective at boosting employment levels (Michalopoulos and Berlin 2000).

In contrast, Canada’s Self-sufficiency Project, which generously supplemented earnings from low-paying full-time jobs, has been very successful at increasing employment and incomes among sole parents who had been long-term beneficiaries (Card and Robbins 1996; Greenwood and Voyer 2000; Michalopoulos et al 2002). The financial incentives associated with this programme were the most generous of those reviewed by Blank et al (1999) and they were also the most effective.

However, Canada’s Earnings Supplement Program, aimed at displaced workers and repeat users of unemployment benefits, was less effective. The programme for repeat users had no impact. This is most likely because 90% of repeat users expected to return to a previous employer and did not want to jeopardise their chances of doing so. Again, this finding highlights the issue of security: higher financial gains were not enough to offset the potential risk of taking up employment that might not work out. The programme had modest effects on displaced workers, who moved into employment – and better employment – more quickly than those in the control group. However, the effects were short-lived.

There is evidence that tax credits have been largely successful at increasing employment levels by increasing the amount of income associated with employment. In one study, the UK’s Family Credit was shown to have reduced employment exits among lone parents that short-term crises, like a sudden drop in earnings, might otherwise have triggered (Bryson and Marsh 1996; McKay and Marsh 1994). Around the time that the Working Families Tax Credit (a more generous tax credit that replaced the Family Credit in 1999) first came into effect, a series of studies estimated that the potential employment effects would be positive and significant (Blundell and Reed 2000). Simulation studies suggested that the Working Families Tax Credit would have a sizeable positive impact on increasing employment among single mothers and among the female partners of unemployed men.

In the US, an expansion of the Earned Income Tax Credit (EITC) has been attributed to an increase in employment among lone mothers (Eissa and Liebman 1996). Meyers and Rosenbaum (2000) estimate that about half of the increase in single mothers’ labour force participation in the US between 1984 and 1996 can be attributed to expansions in the EITC. Ellwood (2000) concludes that the expansion of the EITC, welfare reforms more generally, and the strong economy in the late 1990s led to a “truly unprecedented increase in labour market activity by low-income single parents” (p. 45).

In New Zealand, financial disincentives to work part-time were identified in a study of lone parents in 1993 (Levine, Wyn and Asiasiga 1993). Subsequent policy changes in the benefit abatement rate, allowing for more earnings from part-time work to be kept, appear to have facilitated part-time employment among sole mothers (Wilson 2000; Wilson and Ball 2000).

Although generally positive, the evidence on earnings disregards, earnings supplements and tax credits is mixed. Most likely, this reflects the barriers to employment that beneficiaries face that go beyond financial disincentives. For example, while Minnesota’s earnings disregards programme alone had only minimal effects on employment, these effects were doubled when the programme was combined with employment services, such as childcare and transportation. The effect was greatest among long-term beneficiaries (Michalopoulos and Berlin 2000). As mentioned earlier, Canada’s Self-sufficiency Project had a much greater impact on participants who received job search assistance in addition to earnings supplements. In addition, despite its success compared with the control group, between one-half and two-thirds of programme participants did not take up employment during the year that the supplements were offered. Half of those not taking up employment reported that they could not find a job, one-quarter reported personal or family responsibilities that inhibited their potential employment, and 19% reported poor health or disability.

Together, these findings suggest that financial disincentives often act alongside a range of other barriers to inhibit employment among beneficiaries. The following were particularly important:

for lone parent beneficiaries, the significance attached to the parenting role (Marsh 1997; Puniard and Harrington 1993)
for the long-term unemployed, poor health (Marsh 1997)
for beneficiaries with disabilities, employer discrimination (Hoynes and Moffitt 1995).

It may be, as Corden and Sainsbury (2001) conclude, that incentive measures are mostly influential when they take account of these other factors.

2.5.2 Inadequate social service support and case management

Case managers may be particularly ill-prepared for dealing with clients who have complex situations and multiple barriers to employment, which will inhibit clients’ participation in work-related activities.

Several studies suggest that the nature of case management can have significant impacts on the employment rates of beneficiaries and on employment sustainability. The level of social support provided by case managers is vital. In addition, case managers who have the resources to do their job and who understand their role as a job search adviser are more effective at moving beneficiaries into employment.

In a study of long-term beneficiaries, Sansone (1998) reports that social support resources were critical in beneficiaries’ progress toward independence. The author studied long-term welfare recipients who participated in a state job training programme that provided comprehensive education, job training and support services through intensive case management. The programme design recognised that some barriers to employment among long-term beneficiaries require instrumental aid in the form of childcare or transportation, while others require various types of social support in the form of intensive case management (which is usually missing from programme designs). The results suggest that social support that boosted self-esteem was particularly important in moving beneficiaries closer to employment. Based on research on Work First participants in North Carolina, DeBord et al (2000) support this conclusion and suggest that the importance of social support from involved case managers cannot be overemphasised. Recent studies in the UK have also demonstrated that case managers – or Personal Advisers in the New Deal scheme – can play a critical role in successful job search and the employment of beneficiaries (Lewis et al 2000; Woodfield and Finch 1999).

A key finding of one evaluation of the UK New Deal for Lone Parents was that the role and manner of the Personal Adviser played an important part in shaping participants’ view of the success of the programme (Finch et al 1999). Personal Advisers were judged by participants to be effective in aiding job search if they held personal qualities related to social support, such as friendliness and a positive attitude, rather than if they provided practical assistance. Woodfield and Finch (1999) also found that one of the most important elements identified in the “innovative schemes” to assist lone parents into employment was the opportunity for lone parents to meet each other and provide mutual support. Evans (2000) reports that the information, encouragement and support provided by case managers were identified by participants as the most important elements in helping their transition into employment in the highly successful Bootstrap programme. Two-thirds of participants in that programme found employment, and, overall, 94% took up either employment or some type of training. As a result of the programme, changes in confidence and determination were sustained up to nine years later. Once in employment, social support can aid in job stability. Barnes, Thornton and Campbell (1998) suggest that social support may be particularly important for people with disabilities who have returned to work.

The way in which case managers understand their role and, more broadly, how the programme is marketed are also critical. In Australia, many Department of Social Security (DSS) staff reported seeing their role solely in terms of providing income support for their disabled clients, not in terms of helping them to find work (Jonczyk and Smith 1990). New Zealand case managers were found to be given priority to work-tested clients (DPB beneficiaries with their youngest child over 14), and spent minimal time discussing work preparation options with those who were not required to work right away (DoL and MSD 2001). In this case, the onus was placed on beneficiaries. In addition, there was low staff awareness of the various forms of assistance available to non-work-tested beneficiaries. Greenwood and Voyer (2000) report that part of the success of Canada’s Self-sufficiency Project may lie in the aggressive way it was marketed, with strong messages about making work pay and follow-up phone calls to beneficiaries. In contrast, the less successful Earnings Supplement Program was not marketed as strongly and was represented as a last resort.

Some studies caution that case managers may be ill-equipped to deal with applicants who have multiple, complex and long-term barriers to employment (Brown 2001).

2.5.3 Lack of awareness

There is evidence to suggest that beneficiaries are often unaware of their eligibility for employment services, earnings disregards, and assistance with transition costs that might otherwise boost their employment levels.

As mentioned in Section 2.5.1, one barrier for finding employment may be a lack of knowledge about employment services, earnings disregards and assistance with transition costs. Polit et al (2001) found that beneficiaries typically had little knowledge about the kinds of benefits already available that might ease the financial burdens of the transition to employment. In a review of several US state studies, Schumacher and Greenberg (1999) found limited knowledge of childcare subsidies among eligible former welfare recipients. In Australia, Jonczyk and Smith (1990) found poor knowledge of available services among disabled DSS clients. Over two-thirds did not know there was a service aimed at helping them find employment. In New Zealand, Levine, Wyn and Asiasiga (1993) found that many sole parents lacked information about how the benefit system worked, eg how much they were allowed to earn before benefits were cut, or about available training and employment opportunities. Also in New Zealand, Duncan, Kerekere and Malaulau (1996) found that many lone-mother beneficiaries lacked information about their benefits. A recent evaluation study of benefit reforms for lone parents suggests that many sole parents are unaware of assistance that is now available to them once they leave work (DoL and MSD 2001).

3. Key issues

3.1 Multiple Barriers to Employment

What distinguishes long-term beneficiaries from others is not the presence of any one significant barrier, but the presence of a cluster of barriers and, as will be discussed below, a history of disadvantage or negative life events. The number of welfare recipients in the US with multiple barriers is estimated to be about one-half (within a range of 44–64%) (Patel 2001). Several studies in the US show that the more barriers a welfare recipient has, the lower their likelihood of exiting to or retaining employment (Danziger et al 2000; Fagnoni 2001; Jackson, Tienda and Huang 2001; Loprest and Zedlewski 2002; Scott, London and Edin 2000; Sweeney 2000; US General Accounting Office 2001). In the USGAO review study, 88% of welfare recipients with no barriers were working, compared with 59% with one barrier and 27% with three or more. However, the fact that there is a percentage with multiple barriers who do exit to or retain employment, while some without barriers do not, suggests that there are other characteristics (currently unmeasured or poorly measured) that are affecting employment probability. Patel (2001) reports a high rate of multiple barriers (including low education levels, limited work experience, poor health and disabilities) among welfare families sanctioned for noncompliance with programme rules.

In a UK study, Ashworth, Hartfree and Stephenson (2001) find that individuals who move between the incapacity and unemployment benefits tend to be among the least advantaged, facing multiple barriers including poor health, low levels of qualifications and few job skills. In New Zealand, Lunt and Pernice (1999) point out that people with disabilities tend to face multiple disadvantages and are overrepresented among those who are unemployed and poor.

How do these multiple barriers cluster? One study in Utah characterised beneficiaries’ barriers according to whether they involved physical/mental barriers, work/education barriers or family barriers. They found the most common cluster of barriers (experienced by two-fifths) was a combination of physical/mental barriers and family problems. One-fifth faced barriers in all three areas. Danziger, Kalil and Anderson (2000) grouped barriers into four categories: physical health, mental health, substance dependence and human capital. Physical health, mental health and human capital were found to be most important: women who had barriers in all three of these domains had significantly longer welfare histories than women in almost every other barrier-combination group and than women with none or only one barrier. For example, women with barriers in all three of these domains spent on average 16.5 years since age 18 on welfare, compared with less than nine years among women with barriers in only two of those domains. The authors also provide interesting evidence of how barriers tend to cluster. Mental health barriers tend to occur in isolation, with 18.5% of the sample having only this type of barrier, whereas physical health and substance dependence barriers tend to occur with other barriers.

Dockery and Webster (2001) provide a useful discussion of the problems with identifying individuals at risk of long-term unemployment from a list of individual barriers. The long-term unemployed in Australia, for example, tend to be less educated, less skilled and less proficient in English, and tend to live in lower SES locations and live with other non-working adults. However, few of the long-term unemployed have all of these characteristics, and little is known about whether certain types of long-term unemployed workers, with certain clusters of characteristics, can be identified. Toward that end, the authors perform a cluster analysis of the long-term unemployed and identify five clusters of characteristics that might help with identifying at-risk individuals. However, data are still lacking on the full range of determinants of long-term unemployment, and the authors suggest that more information is needed about the social and health histories of the unemployed, such as information about past experience with domestic violence, drug dependence, sexual abuse, homelessness, job instability and criminality. Likewise, Athanasou, Pithers and Petoumenos (1995), also in Australia, caution that, while the usual social/ demographic risk factors do distinguish between the long-term unemployed and the very long-term unemployed at the aggregate level, they are poor predictors at the individual level.

3.2 Employment Sustainability

As Harris (1993; 1996) and Pavetti (1997) have found, the employment problem for beneficiaries may be less about finding employment than about keeping it. Many studies document the common pattern of cycling on and off benefit (Gobbi and Rea 2002; Moffitt 2001), and a large body of evaluation literature on welfare-to-work programmes suggests that beneficiaries who leave welfare for work are highly likely to leave employment within a year (Friedlander and Burtless 1995; Hershey and Pavetti 1997; Parker 1994; Quint, Musick and Ladner 1994; Riccio and Freedman 1995; Strawn and Martinson 2000; Strawn, Greeberg and Savner 2001). In the US, research suggests that the situation may have improved in recent years (Holzer, Stoll and Wissoker 2001; Johnson 2001), although employment sustainability is likely to be affected by the overall health of the economy (see Section 2.4.2). White and Forth (1998) report that three out of four unemployed people who take up employment are underemployed or self-employed, or have jobs that are temporary and/or part-time. One-third of all unemployment beneficiaries (before the Job Seeker’s Allowance was introduced in the UK) who left the benefit returned within 30 weeks (McKay, Walker and Youngs 1997).

Harris (1996) found that the welfare leavers who remained in employment were those with more than a high school education, with prior work experience and with fewer than three children. However, within these groups, there was still wide variation in who remained employed and who returned to welfare. As Kalil et al (1998) point out, this is most likely the result of unmeasured characteristics, along the lines of the many and multi-leveled barriers described above. More recently, Polit et al (2001) demonstrate that multiple barriers are associated with employment instability among beneficiaries who left welfare for employment. Nearly all who returned to welfare after a spell of employment had personal or family barriers, whereas the most stably employed faced the fewest (although the majority had at least one).

In a study of Florida welfare leavers described by Schumacher and Greenberg (1999):

about one-quarter said that they knew of a better job they could get if they had appropriate childcare
over one-fifth had been absent from their job during the previous months because of a childcare problem
one-third reported that childcare problems had led them to change their working hours
over one-fifth had changed jobs because of childcare problems.

Overall, half the survey respondents identified at least one of these childcare-related events. Debord et al (2000) found that welfare recipients were more likely to drop out of a job training programme if they were dissatisfied with their childcare provider, had concerns about their children’s safety, or if the childcare site was located some distance from home.

Importantly, though, these personal and family barriers are reinforced by the institutional context. In a study by Polit et al (2001), those who returned to welfare tended to leave jobs that were low-paying, full-time jobs with no fringe benefits, such as paid sick or vacation days: 77% of these jobs had no such benefits, and fewer than half had employer-provided health insurance, essential in the US. In addition, less than one-half of those who subsequently applied for unemployment insurance received it because their previous employment experience failed to meet eligibility requirements. In Australia, too, McHugh and Millar (1996) cite a study of JET clients who returned to benefit receipt after a spell of employment (Zanetti 1994). This study found that these beneficiaries had a difficult time remaining in jobs that lacked family-friendly workplace provisions, like the ability to take paid leave to care for a sick child.

Holzer, Stoll and Wissoker (2001) call attention to this “match” between the needs of the individual and the employer. They found that over 40% of the 750 employers surveyed complained about absenteeism among employees who were former welfare recipients. In addition to absenteeism, employers reported that poor attitudes toward work and poor relations with co-workers were more problematic than welfare recipients’ low levels of basic skills. This finding supports the view the “soft skills” (Moss and Tilly 2001) may be key barriers to sustainable employment among beneficiaries. However, Holzer, Stoll and Wissoker (2001) also find that the major causes of absenteeism were childcare and transportation problems. The authors suggest that “a combination of ‘hard’ and ‘soft’ (or cognitive and non-cognitive) skills, along with ‘matches’ to appropriate jobs, are necessary for even minimal success in the labor market among these recipients” (p. 7). A good match might involve proximity to public transportation for those without access to a car, or to on-site childcare. Their survey of employers also showed that the few employers who provided childcare and transportation assistance to employees experienced less absenteeism.

Studies in the UK suggest that similar mismatches between the needs of employees and employers may be involved in job instability among lone parents. Walker and Kellard (2001) report that a significant number of jobs taken by formerly inactive lone parents may end because of a dispute or disagreement between the employee and the employer. Qualitative interviews with lone parents who were part of the Personal Adviser scheme revealed two main “causes” of job exits: inappropriate match of the work or work environment to the individual, and employer attitudes and practices (such as changing work hours) (Lewis 2001). Other problems identified in sustaining employment were: increased living expenses associated with employment, breakdowns in childcare arrangements, and the difficulties of balancing parental and employee responsibilities.

Finally, a recent review of research on the predictors of employment retention among parents leaving welfare for work in the US (Patel 2001) documented the following key factors: starting off in a higher-paying job (Rangarajan, Schochet and Chu 1998; Rangarajan, Meckstroth and Novak 1998), working steadily initially (Cancian and Meyer 2000), finding jobs with employer benefits, like paid vacation (Rangarajan, Meckstroth and Novak 1998), and working in non-sales occupations (Cancian and Meyer 2000). Factors that predict job advancement among these welfare leavers include: starting off in a higher-paying job (Cancian and Meyer 2000), having or acquiring higher basic skills, education or training (Cancian and Meyer 2000; Corcoran and Loeb 1999), and starting off in non-sales occupations (Cancian and Meyer 2000). Again, the mix between the individual factors (higher levels of education and skills, the ability to get a higher-paying job, possibly prior experience in certain occupations) and structural factors (employers who pay well and provide benefits) plays a key role.

3.3 Detrimental Effects of Long Spells On-Benefit

One of the key debates among welfare researchers is whether welfare provision encourages dependency – is long-term welfare receipt itself a barrier to finding employment? This process is termed “state dependence” or “duration dependence” and is contrasted with “heterogeneity” (Gottschalk and Moffitt 1994). The theory of heterogeneity suggests that the reason long-term recipients have an increasingly difficult time leaving welfare is because those who are better off in terms of skills, education level and the like leave welfare earlier. This “leaves behind” a population of long-term recipients who have especially disadvantageous labour market characteristics.

Although efforts to empirically demonstrate the nature and magnitude of duration dependence have been limited by data constraints and unmeasured or unmeasurable characteristics that might influence long spells on welfare, the available evidence suggests that duration dependence does exist. While Chapman (1994) says that results from Australian research on the subject have been ambiguous, UK research provides evidence of duration dependence, indicating that early intervention is required for some. Again in the UK, Sheehan and Tomlinson (1998) find evidence of a strong negative duration dependence in a study of the long-term unemployed. Using monthly US administrative data, Chay, Hoynes and Hyslop (1999) provide evidence for the existence of duration dependence at the individual level. Sandefur and Cook (1998) found that the longer one stays on welfare, the lower the chances of leaving permanently, even after controlling for individual and family characteristics. In addition, the authors find that chances of permanently exiting from welfare depend on a number of other characteristics, such as number of children, qualifications and job availability.

What might duration dependence signify? Several of the studies looking at duration dependence among the unemployed suggest that the two major factors are skill deterioration and “negative signals” sent to employers about the employability and productivity of the individual. Chapman (1994), for example, suggests that empirical evidence from previous research has established that prior interaction with “mainstream” employment positively affects job search. Employers use statistical discrimination in a sense, using long-term unemployment as a proxy for productivity in the absence of any other information. In order to prove them wrong, the long-term unemployed would need access to employment. Sheehan and Tomlinson (1998) also suggest that the evidence supports the idea that employers use long-term unemployment as a negative signal about a potential employee. Sandefur and Cook (1997) suggest that the long-term welfare population probably represents a mix of those who were severely disadvantaged in the labour market to start with and those who had the skills and characteristics necessary for employment to start with but who have since experienced skill deterioration and decreased attractiveness to employers. The challenge, then, is to provide a safety net for those who will always be severely disadvantaged in the labour market, while providing early intervention for those for whom employment is viable (Sandefur and Cook 1997).

Social-psychological research provides insights into other mechanisms that might contribute to duration dependence. Research in the UK has found a link between economic hardship and low morale, such that spending long periods of time in severe hardship creates a barrier to employment by depressing the morale needed for job search (Finlayson et al 2000; Marsh et al 2001). Millar and Ridge (2001) point out that a selection effect may be operating, such that those with better morale and less hardship move into work more quickly, leaving behind a population on-benefit in the long term who have low morale to start with and then get further discouraged as time goes by. In addition, Coiro (2001) in the US finds an association between persistent welfare dependence and greater risk for depression, which has been identified as a potential barrier to effective job search and stability. Other research indicates that those on benefits and detached from the labour market for long periods of time show a decline in physiological and psychological functioning (Wadsworth, Bartley and Montgomery 1999). Thus, both physical and mental health may be affected adversely.

For a large proportion of the barriers identified in this report, research has pointed to a process of reciprocal causation. This is where a particular barrier might play a role in inhibiting employment but non-employment plays a role in creating or worsening that and other barriers. For example, this process has been identified in terms of mental illness and psychological distress (Coiro 2001; Danziger, Carlson and Henly 2001), alcohol misuse (Khan, Murray and Barnes 2002), literacy (employment serves to maintain or enhance literacy) (Kapsalis 1998) and transportation barriers (Social Exclusion Unit 2003). This entrenchment of barriers may be a major factor in non-employment of long-term beneficiaries. However, employment per se may not be the answer for preventing the development or exacerbation of these barriers – instead, employment that pays above poverty-level wages may be required (Dickens and Ellwood 2001; Dooley and Prause 1998; Pettersen and Friel 2001; Strandh 2001).

3.4 Accumulation of Negative Life Events

The role of the life course – ie the timing, sequence and duration of life events prior to a welfare spell – should not be overlooked. Often barriers are the result of an accumulation of negative life events, starting in childhood. Human capital characteristics (primarily cognitive skills) prior to entering the labour market as a young adult are associated with early job instability, which in turn appears to play an independent role in affecting later-life job instability (Holzer and LaLonde 1999). Entering the labour market at a period of high unemployment appears to have a lasting negative impact on the employment probabilities of the less skilled (Burgess et al 1999). The effect on the later-life employment probabilities of the highly skilled is actually positive, possibly because adverse labour market conditions at age of entry encouraged many of the highly skilled to pursue further education. Several studies implicate teenage motherhood as a major risk factor for long-term benefit receipt due to its effects on educational attainment, work experience, ultimate family size and, possibly, relationship history (Bradshaw and Millar 1991; Harris 1993; Klepinger, Lundberg and Plotnick 1997; Leahy, Buss and Quane 1995; Pavettii 1999).

Using the Christchurch Health and Development Study, Seth-Purdie (2000) presents striking evidence for the influence of negative childhood events and circumstances on later welfare risks and dependence. Numerous adverse characteristics were measured, including exposure to childhood sexual abuse, parental alcoholism and criminality, and regular or severe physical punishment. Over one-fifth of women who were receiving welfare at age 21 were exposed to one or more of these conditions. Those who at 21 had already experienced two or more years of unemployment were much more likely to have experienced or participated in problematic teenage behaviours, like frequent truancy or alcohol and drug abuse. As with multiple barriers to employment, the greater the childhood adversity experienced, the less likely the woman was to have never experienced unemployment or long-term unemployment. The author states: “[M]any of the people who are most expensive in terms of health, criminal justice, and welfare services, are individuals on whom adverse circumstances have literally left their mark in the form of human capital deficits including a vulnerability to stress and heightened risk of chronic health problems in middle age” (Seth-Purdie 2000, 51).

3.5 Understanding and Addressing Barriers

A comprehensive understanding of long-term beneficiaries’ employment barriers requires attention to personal, family, community and institutional factors, as well as their interactions. A comprehensive approach to addressing these barriers requires a mix of services and individualised case management.

As this review has shown, a beneficiary might face numerous potential barriers to employment. Indeed, beneficiaries who face multiple barriers are most likely to remain on-benefit in the long term. Because of this complexity, it is difficult to identify who will become long-term beneficiaries at the beginning of a spell of benefit use without a wide range of information about the individual’s current situation and past history.

Findings from research on long-term beneficiaries and the policies aimed at helping them suggest two things. First, a multi-level approach that takes into account individual, family and structural factors would be the most fruitful. Marcenko and Fagan (1996:116) advocate the “person-in-environment” approach:

[The person-in-environment approach] focuses on the multiple levels of social systems which are influenced by or have an influence on individuals. These include the family, neighborhood, local, state, and national systems. Each of these systems, in turn, has the potential to provide resources and generate stress for policy and programme recipients. In order to evaluate the potential efficacy of a policy, it is necessary to assess the impact, both negative and positive, that the policy is likely to have on the person and her environment.

The second is that a life course approach that taps into longitudinal sources of data is necessary. Research consistently shows that beneficiaries with the most barriers have an accumulation of disadvantages, many of which have taken hold in childhood. Personal history matters.

The policy-development corollaries to these research approaches are prevention, and flexible, individualised case management that recognises and addresses multi-faceted barriers.

Although the aim of this report was not to review in depth the results of evaluation studies of welfare-to-work programs (see Hand 2001 for a review), several key pieces of evidence on barriers to employment have come from evaluation studies aimed at addressing clusters of barriers. Taken together, these studies emphasise that flexible, “mixed” or “balanced” approaches, which offer a range of services to meet the diverse barriers of beneficiaries, are most effective (US Department of Health and Human Services and US Department of Education 2001a&b). Additionally, an intensive, individualised approach by case managers (who act as “coaches”) could result in long-lasting effects (DeBord et al 2000; Evans 2000; Lewis et al 2000; Sansone 1998; Woodfield and Finch 1999).

4. The New Zealand evidence

Results from New Zealand research on barriers to employment among beneficiaries were interspersed in the general review above. Here the focus is on the long-term beneficiary population specifically and future research that is needed to fill gaps in our knowledge.

4.1 The Long-Term Beneficiary Population

It is only recently that the population of long-term beneficiaries could be studied in New Zealand. MSD’s benefits dynamics data set (Wilson 2001) and the Department of Labour’s database of job seekers registered with the New Zealand Employment Service (Gobbi and Rea 2002) make it possible to identify long-term benefit recipients, including those who stay on-benefit for a long continuous spell and those who are frequent (but not necessarily long-spell) recipients.

Looking at working-age beneficiaries who started a spell of benefit receipt in 1993, Wilson (2001) found that long, continuous spells were rarer than might be assumed using cross-sectional data. Only one-fifth of these spells lasted for a year or more. However, counting all spells over a five-year period, almost two-thirds spent at least one year in receipt of benefit income, and one-third spent three or more years on benefit income. Studies of unemployment beneficiaries in New Zealand have also shown that multiple spells are common (De Radd 1997; Gardiner et al 1994; Gobbi and Rea 2002). Gobbi and Rea (2002) found that most registered job seekers who end an unemployment spell register as unemployed job seekers again within the next four years, typically within a year and a half. About one-quarter of all job seekers could be considered chronically unemployed, experiencing a median of three distinct unemployment spells and, on average, spending two and a half years out of four unemployed.

Thus, long-term beneficiaries are characterised mainly by repeated spells of benefit receipt that lead to a lengthy amount of time on-benefit over a given time period. A smaller proportion experience long continuous spells. Wilson (2001) finds differences in the characteristics of these groups, suggesting that they may be distinct populations with different barriers (Moffitt [2001] makes a similar suggestion based on US data). For example, those with many shorter spells tend to be young and without dependants, whereas those with long first spells or a lengthy total duration are more likely to have children, especially young ones.

4.2 Individual Barriers to Sustainable Employment among Beneficiaries

As Wilson (2001) suggests, it is most likely the total length of time a beneficiary spends on-benefit that is of concern to policy makers, rather than the continuity of their spells. Therefore, the finding that long-term benefit receipt is most often characterised by repeated spells on-benefit highlights the importance of employment sustainability, and not just the employment transition, for New Zealand policy makers.

What are the barriers to sustainable employment among beneficiaries in New Zealand? Generally speaking, evidence comes from three sources:

large-scale, representative administrative data sets, such as those described above, that are comprehensive in coverage but limited in terms of identifying barriers
small-scale, non-representative samples of beneficiaries that provide rich information on the nature of employment barriers
anecdotal evidence.

The most useful data sets for identifying the long-term beneficiary population (those described above) are very limited in terms of identifying barriers to employment. Analyses are restricted to the demographic data collected. More information on barriers identified by beneficiaries is available from smaller-scale studies but these rarely focus specifically on the long-term benefit population. The strongest evidence base exists, then, for those barriers that have been identified by more than one method. Most studies focus on one specific beneficiary type rather than all beneficiaries at once.

4.2.1 Lone parents

Patterns of long-term receipt among beneficiaries with young and/or numerous children identified by the benefits dynamics data set provide evidence that childcare is a barrier, but which aspects of childcare are problematic (eg availability, costs, concerns about non-parental care) cannot be identified with this data set. However, three detailed studies of DPB recipients provide further evidence (DoL and MSD 2001; Levine, Wyn and Asiasiga 1993; Wylie 1980). Echoing findings from the earlier study by Wylie (1980), Levine, Wyn and Asiasiga (1993) found that a large proportion (about half) of sole mothers preferred to stay home while their children were young, and that those using childcare preferred informal care by friends or relatives. This is consistent with Wilson’s (2001) finding that, among DPB recipients with a very young child, the average total duration on-benefit was lower for Pacific peoples than for Päkehä. Wilson suggests this may be because Pacific women with young children have higher rates of full-time employment, possibly because of extended family structures that facilitate mothers’ employment by providing access to informal childcare.

In the study by Levine, Wyn and Asiasiga (1993), among those mentioning childcare as a problem, the cost of childcare was most problematic, with some beneficiaries deciding to wait until their youngest child reached school age in order to reduce the costs of childcare. In addition, sole mothers who worked non-standard hours and those who did not have a car or did not live near a childcare facility mentioned availability as a problem. Supporting these findings, a government report on the effect of the 1999 policy reforms on DPB recipients shows that the accessibility and cost of childcare were crucial factors affecting employment sustainability for those with a youngest child under the age of 14 (DoL and MSD 2001). In both studies, some beneficiaries with children aged 14 and older also considered childcare to be a problem because they felt that their teenagers required adult supervision.

Another common finding was the financial disincentive for working part-time as a result of the benefit abatement rate. Using the benefits dynamics data set, Wilson (2000) and Wilson and Ball (2000) provide strong evidence that the 1996 policy reforms aimed at reducing these disincentives had a significant effect on increasing part-time employment among beneficiaries. The evidence base for the 1996 policy reform came from the Levine, Wyn and Asiasiga (1993) study, which documented the problems that sole mothers faced in taking up their preferred type of employment.

More recent post-reform evidence suggests that some sole mothers still face financial disincentives for part-time work (DoL and MSD 2001). In the government study of the 1999 reforms, the researchers found that sole mothers who took up part-time employment earned lower hourly wages than those who moved into full-time work. In addition, part-time workers noted that the financial value of working part-time, once the transition and ongoing costs of employment were taken into account, was only marginal.

Other barriers to employment among lone-parent beneficiaries documented include: lack of effective case management; the transition costs of moving into work; debt and material hardship; low skills and qualifications; lack of access to “family friendly” employers; lack of information about employment assistance; perceived employer discrimination; and anxiety and lack of confidence (DoL and MSD 2001; Levine, Wyn and Asiasiga 1993; Wylie 1980). Recently, a suite of studies has documented the importance of health (of lone parents themselves and of their children) as a multi-dimensional barrier to employment among DPB recipients (Baker 2002; Baker and Tippin 2003).

4.2.2 The Long-term Unemployed

As with lone parents, the administrative data sets provide information on the demographic characteristics of the long-term unemployed. Wilson (2001), looking at all benefit types, found that younger and older recipients are both more likely to be on-benefit in the long term. Mäori are more likely than Päkehä to have a long total duration. These ethnic differences are particularly pronounced for those under 30. Wilson suggests that this reflects differences in the nature of youth unemployment among the two groups, with young Päkehä more likely to have received the Emergency Unemployment Benefit while in tertiary education – a short-term stop-gap measure. Gobbi and Rea (2002) also found that young people, sole and mixed Mäori, and Pacific peoples spent the greatest total time registered unemployed. Mature workers over 50 were more likely to have a long single spell of registered unemployment. Individuals with no formal qualifications were more likely to have long single spells and multiple spells, and to spend the greatest total time registered unemployed. The authors also found that individuals from particular regions were more at risk of long-term unemployment (variously defined), with those from Northland disadvantaged in each measure of unemployment duration.

What employment-barrier mechanisms lie behind these patterns? As Watson, Mare and Gardiner (1997) point out, measurement of the actual barriers to employment, rather than just demographic characteristics, is necessary in order to identify early on those at risk of long-term unemployment. The authors find too much variation within each at-risk demographic group for the measures to be of any use in predicting who will become unemployed at the individual level. The authors conclude that data on psychological and social variables are needed, as well as more detailed information on local labour market conditions.

The most useful recent study for identifying some of these barriers is the nationwide survey of very long-term job seekers – over 52 weeks of continuous registration with the New Zealand Employment Service (Parker 1997). The study identified limited mobility as an especially important barrier, and documented the way that the nature of mobility as a barrier varies according to region. For those living in Gisborne, for example, the main problem was a lack of driver licence, whereas a high proportion of those living in Taranaki reported lack of access to public transportation. This highlights the dual nature of transportation as a barrier at both an individual and community level. The study also found that Mäori were particularly likely to report mobility problems of all types.

Over half of the respondents said that the uncertainty of taking up employment and potential income instability if employment did not work out were significant barriers to employment. The percentage among Mäori was higher (60%). Fewer reported that they did not think they could earn as much in the labour market as they received in benefit income (36%). Again, percentages were higher among Mäori (41%) and Pacific peoples (46%).

Those unemployed for more than two years (77%) reported disincentives of the benefit system. A lack of qualifications was perceived to be a barrier (67%), as was overall discouragement (63%) and a lack of available jobs in the local area (53%). Nearly half (46%) said that they faced unaffordable work-related costs such as work clothes. Perceived employer discrimination on the basis of age was especially high among older job seekers, with 80% reporting this as a barrier. Some Pacific job seekers (33%) and those of “Other” ethnic identity (39%) reported poor English language skills as a barrier. Health problems affected 36% of the total sample and 42% of those who had been unemployed for over two years. Parker concludes that, overall, limited mobility, perceived age discrimination, and a lack of qualifications were the three main barriers to employment. Anecdotal evidence also suggests that age discrimination is a major barrier among the mature-age unemployed (Patterson 1999).

4.2.3 People with disabilities

Research is needed on employment barriers among people with disabilities who are long-term beneficiaries in New Zealand.

4.2.4 Missing evidence

Some of the individual barriers found in the international literature either have not yet been identified in the New Zealand context (presumably because they are not a problem or have not yet been measured properly) or have been found to be not especially significant.

Given the prominence of health issues as one of the major barriers to employment in the international literature, it is surprising that health problems have not (until very recently) been given more attention in the New Zealand context. Results from a recent health survey of DPB recipients in Otara, Brown’s Bay and Kaitaia and a related qualitative study (Baker 2002; Baker and Tippin 2003) suggest that health problems may be significant among the DPB population, although more research will be needed in other locations to confirm these findings. Health problems were mentioned by long-term job seekers in Parker’s (1997) study, but they did not rate as one of the most significant barriers.

More research is also needed on: mental health issues, substance dependence, domestic violence, social networks, social support, and local labour market conditions (for beneficiaries specifically). In addition, more research is needed specifically on people with disabilities, discouraged workers (those who are not employed and not registered as job seekers), and long-term DPB recipients.

4.3 Multiple Barriers

Most prominently missing from the New Zealand knowledge base is evidence on multiple barriers to employment among beneficiaries. The international literature review suggests that long-term beneficiaries are characterised by facing multiple barriers to employment, possibly at various contextual levels. The most useful study yet is Seth-Purdie’s (2000) analysis of the Christchurch Health and Development Study, results from which are discussed at some length in Section 3.4. More research is needed to understand how employment barriers tend to cluster among long-term beneficiaries and how these barriers are formed across the life course. Findings from the Christchurch Health and Development Study suggest that, at least for some of the most disadvantaged beneficiaries, the clustering of multiple barriers is a process that takes hold in childhood. The cumulative effect of these negative life events can result in multiple and compounding disadvantages as the adult life course proceeds. More research is needed to understand this process. Both retrospective and prospective approaches could be taken, in the first instance, by surveying long-term beneficiaries to understand the opportunities and constraints they have faced along the life course, and, in the second, by using panel data to predict spells of long-term receipt based on prior history.

Caution should be exercised when trying to develop and use risk profiles as a means of identifying those at risk of long-term benefit receipt. Several studies suggest that prediction at the individual level is poor (Athanasou, Pithers and Petoumenos 1995; Dockery and Webster 2001; Watson, Mare and Gardiner 1997) and that clients with multiple barriers require one-on-one, individualised, intensive case management in order to identify the particular array of barriers that each individual faces (Sandefur and Cook 1997; Seth-Purdie 2000; Brown 2001). One has only to look at the long list of barriers presented in this document to appreciate the potential complexity of developing a “one-size-fits-all” approach to identifying at-risk individuals. Still, it would be useful to look at each barrier and then at clusters of variables and examine how much variation there is in benefit patterns among each of these subgroups.

Finally, more research is needed to understand the interplay of both barriers and “bridges” to employment. What personal and contextual factors facilitate employment, even in the presence of multiple barriers? The wide variation in work behaviour found among at-risk individuals most likely reflects such differences. A good example is social support. While a lack of social support can be named as a barrier to employment, it is perhaps more accurate to name the presence of social support as a bridge to employment that helps individuals overcome the barriers they face. Thus, a key factor distinguishing benefit use patterns among those facing the same cluster of barriers could be access to social support. This is an area that deserves further research, not only in New Zealand but also in the other countries reviewed here.

5. References

Abt Associates Inc. (1999) The ABC evaluation: Do welfare recipients’ children have a school attendance problem? Cambridge, Massachusetts.

Adler, M. (1993) Disability among women on AFDC: An issue revisited, Report prepared for the Office of the Assistant Secretary for Planning and Evaluation, US Department of Health and Human Services, Washington, DC.

Ahluwalia, S.K., S.M. McGroder, M.J. Zaslow, and E.C. Hair (2001) Symptoms of depression among welfare recipients: A concern for two generations, Research Brief, Child Trends, Washington, DC.

Anderson, P.M and P.B. Levine (2000) Childcare and mothers’ employment decisions, in D. Card and R.M. Blank (eds), Finding jobs: Work and welfare reform, Russell Sage Foundation, New York.

Arthur, S., A. Corden, A. Green, J. Lewis, J. Loumidis, R. Sainsbury, B. Stafford, P. Thornton, and R. Walker (1999) New Deal for disabled people: Early implementation, Research Report No 106, (UK) Department of Work and Pensions, Corporate Document Services, Leeds.

Ashworth, K., Y. Hartfree, and A. Stephenson (2001) Well enough to work? Research Report No. 145, (UK) Department of Work and Pensions, Corporate Document Services, Leeds.

Athanasou, J.A., R.T. Pithers and K. Petoumenos (1995) Characteristics of long-term unemployed and very long-term unemployed clients: A study of employment service registrants in Australia, Australian Bulletin of Labour 21: 198–207.

Avenilla, F. and S. Singley (2001) Neighborhood effects on child and adolescent development: Assessing today’s knowledge for tomorrow’s villages, in A. Booth and A.C. Crouter (eds), Does it take a village? Community effects on children, adolescents, and families, Mahwah, NJ: Lawrence Erlbaum Associates, pp. 229–243.

Baker, M. (2002) Child poverty, maternal health and social benefits, Current Sociology 50(6): 823–838.

Baker, M. and D. Tippin (2003) More than just another obstacle: Health, Domestic Purposes Beneficiaries, and the transition to paid work, paper presented at the Social Policy, Research and Evaluation conference Connecting Policy, Research and Practice, 29–30 April, Wellington.

Barnes, H., P. Thornton and S.M. Campbell (1998) Disabled people and employment: A review of research and development work, Bristol: The Policy Press and the Joseph Rowntree Foundation.

Barusch, A.S., M.J. Taylor, and S. Abu-Bader (1999) Understanding families with multiple barriers to self sufficiency, Final report submitted to Utah Department of Workforce Services, Social Research Institute, Graduate School of Social Work, University of Utah, Salt Lake City.

Bassuk, E.L., L.F. Weinreb, J.C. Buckner, A. Browne, A. Salomon, and S.S. Bassuk (1996) The characteristics and needs of sheltered homeless and low-income housed mothers, Journal of the American Medial Association 276: 640–646.

Berthoud, R. (1998) Disability benefits: A review of the issues and options for reform, Joseph Rowntree Foundation, York.

Berthoud, R., J. Lakey, and S. McKay (1993) The economic problems of disabled people, Policy Studies Institute, London.

Bittman, M. (2001) Mature age workers – redundancy, ‘early’ retirement and prospective welfare expenditure, Newsletter No. 80, Social Policy Research Centre, University of New South Wales, Sydney.

Bjorklund, A. and T. Eriksson (1998) Unemployment and mental health: Evidence from research in the Nordic countries, Scandinavian Journal of Social Welfare, 7(3): 219–235.

Blank, R.M. (2000) Fighting poverty: Lessons from recent US history, Journal of Economic Perspectives 14(2): 3–19.

Blank, R.M., D. Card, and P.K. Robins (1999) Financial incentives for increasing work and income among low-income families, paper prepared for the Joint Center for Poverty Research Conference Labor Markets and Less Skilled Workers, 5–6 November, Washington, DC.

Blau, D. and E. Tekin (2001) The determinants and consequences of child care subsidies, Poverty Research News 5(3): 3–4.

Blundell, R. (1994) Work incentives and labour supply in the UK, in A. Bryson and S. McKay (eds) Is it worth working? Policy Studies Institute, London.

Blundell, R. and H. Reed (2000) The employment effects of the Working Families Tax Credit, Briefing Note No. 6, The Institute of Fiscal Studies, London.

Booth, A. and A.C. Crouter (eds) (2001) Does it take a village? Community effects on children, adolescents, and families, Lawrence Erlbaum Associates, Mahwah, NJ.

Born, C.E., P.J. Caudill, C. Spera, and J.F. Kunz. (1998) A look at life after welfare, Public Welfare, winter, pp. 32–37.

Bottomley, D., S. McKay, and R. Walker (1997) Unemployment and job seeking, Research Report No. 62, (UK) Department of Work and Pensions, Corporate Document Services, Leeds.

Bradbury, B., K. Norris and D. Abello (2001) Socio-economic disadvantage and the prevalence of disability, Social Policy Research Centre Report 1/01, University of New South Wales, Sydney.

Bradshaw, J. and J. Millar (1991) Lone-parent families in the UK, Research Report No. 6, (UK) Department of Work and Pensions, Corporate Document Services, Leeds.

Brandon, P.D. and D.P. Hogan (2002) The effects of children with disabilities on mothers’ exit from welfare, paper presented at the Joint Center for Research on Poverty conference The Hard-to-Employ and Welfare Reform, 28 February to 1 March, Washington, DC.

Bray, J.W., G.A. Zarkin, M.L. Dennis, and M.T. French (2000) Symptoms of dependence, multiple substance use, and labor market outcomes, American Journal of Drug and Alcohol Abuse 26(1): 77–95.

Breslin, F.C. and C. Mustard (2003) Factors influencing the impact of unemployment on mental health among young and older adults in a longitudinal, population-based survey, Scandinavian Journal of Work Environment and Health 29(1): 5–14.

Brooks, M.G. and J.C. Buckner (1996) Work and welfare: Job histories, barriers to employment, and predictors of work among low-income single mothers, American Journal of Orthopsychiatry 66(4): 526–537.

Brosnan, P., and M. Wilson (1989) The historical structuring of the New Zealand labour market, Working Paper No. 4/89, Industrial Relations Centre, Victoria University, Wellington.

Brosnan, P., M. Wilson, and D. Wong (1989) Welfare benefits and labour supply: a review of the empirical evidence, New Zealand Journal of Industrial Relations 14: 17–35.

Brown, A. (2001) Beyond work first: How to help hard-to-employ individuals get jobs and success in the workforce, Manpower Demonstration Research Corporation, New York.

Browne, A., A. Salomon and S.S. Bassuk (1999) Impact of recent partner violence on poor women’s capacity to maintain work, Violence against women 5(4): 393–426.

Brush, L. (1999) Woman battering and welfare reform: The view from a welfare-to-work program, Journal of Sociology and Social Welfare 26(3): 49–60.

Bryant, R.R., A. Jayawardhana, V. A. Samaranayake, and A. Wilhite (1996) The impact of alcohol and drug use on employment: A labor market study using the National Longitudinal Survey of Youth, Discussion Paper No. 1092–96, Institute for Research on Poverty, University of Wisconsin-Madison, Madison, WI.

Bryson, A., R. Ford, and M. White (1997) Making work pay: Lone mothers, employment and well-being, Joseph Rowntree Foundation, York.

Bryson, A. and A. Marsh (1996) Leaving family credit, Research Report No. 48, (UK) Department of Work and Pensions, Corporate Document Services, Leeds.

Bunt, K., J. Shury, and D. Vivian (2001) Recruiting benefit claimant, Research Report No. 150, (UK) Department of Work and Pensions, Corporate Document Services, Leeds.

Burgess, S., C. Propper, H. Rees, and A. Shearer (1999) The Class of ’81: The effects of early-career unemployment on subsequent unemployment experiences, Centre for Analysis of Social Exclusion Paper No. 32, London School of Economics, London.

Callister, P. (1998) Some geographic dimensions of being work-rich and work-poor: Changes between 1986 and 1996, Social Policy Journal of New Zealand 11: 131–152.

Campbell, N. (1999) The decline of employment among older people in Britain, Centre for Analysis of Social Exclusions Paper No. 19, London School of Economics, London.

Cancian, M. and D.R. Meyer (2000) Work after welfare: Women’s work effort, occupation, and economic well-being, Social Work Research 24(2): 69–86.

Card, D. and P.K. Robins (1996) Do financial incentives encourage welfare recipients to work? Initial 18-month findings from the Self-Sufficiency Project, Social Research and Demonstration Corporation,Vancouver.

Carling, K., B. Holmlund and A. Vejsiu (2001) Do benefit cuts boost job finding? Swedish evidence from the 1990s, The Economic Journal 111: 766–790.

Cervero, R., O.S. Sandoval and J. Landis (2002) Transportation as a stimulus of welfare-to-work: Private versus public mobility, Journal of Planning Education and Research 22(1): 50–63.

Chapman, B. (1994) Long term unemployment: the case for policy reform, Social Security Journal, June, 19–39.

Chay, K.Y., H. Hoynes, and D. Hyslop (1999) A Non-experimental analysis of ‘true’ state dependence in monthly welfare participation sequences, Paper presented at the 1999 Joint Statistical Meetings, University of California, Berkeley, California.

Claussen, B. (1999) Health and re-employment in a five-year follow-up of long-term unemployed, Scandinavian Journal of Public Health 27(2): 94–100.

Cobb-Clark, D., A. Liu, and D. Mitchell (1999) Reassessing the role of child care costs in the work and care decisions of Australian families, Centre for Economic Policy Research Discussion Paper 409, Australian National University, Canberra.

Coiro, M.J. (2001) Depressive symptoms among women reduce likelihood of successfully moving off welfare, in M.C. Lennon (ed), Welfare, work and well-being, Haworth Medical Press, New York.

Coley, R.L., P.L. Chase-Lansdale, and C.P. Li-Grining. (2001) Child care in the era of welfare reform: Quality, choices, and preferences, Policy Brief 01–4, Welfare, Children, and Families: A three-city study, Johns Hopkins University, Baltimore, Maryland.

Colledge, M. (1991) Workforce barriers for sole mothers in Australia, Social Security Journal, August, pp. 10–25.

Colten, M.E., C. Cosenza and M.A. Allard (1996) Domestic violence among Massachusetts AFDC recipients: Preliminary results, Center for Survey Research, University of Massachusetts, Boston.

Comino, E.J., E. Harris, T. Chey, V. Manicavasagar, J. P. Wall, G.P. Davies, and M.F. Harris (2003) Relationship between mental health disorders and unemployment status in Australian adults, Australian and New Zealand Journal of Psychiatry 37(2): 230–235.

Corcoran, M. and S. Loeb (1999) Will wages grow with experience for welfare mothers? Focus 20(2), Institute for Research on Poverty, University of Wisconsin-Madison, Madison, WI.

Corden, A. and R. Sainsbury (2001) Incapacity benefits and work incentives, Research Report No. 141, (UK) Department of Work and Pensions, Corporate Document Services, Leeds.

Coulton, C., M. Su, S. Guo, and N. Bania (1997) Characteristics of new entrants predictive of length of welfare participation, Briefing Report No. 9705, Center on Urban Poverty and Social Change, Case Western Reserve University, Cleveland, Ohio.

Council of Economic Advisers (1999) Technical report: Economic expansion, welfare reform, and the decline of welfare caseloads: An update, Executive Office of the President, Washington DC:

Crowell, L.F. (2001) Welfare reform: Reforming welfare or reforming families? Families in Society: The Journal of Contemporary Human Services, 82(2): 157–164.

Curcio, W. (1996) The Passaic County study of AFDC recipients in a welfare-to-work program: A preliminary analysis, Passaic County (New Jersey) Family Development Program.

Currie, S., K. Foley, S. Schwartz, and M. Taylor-Lewis (2001) BladeRunners and Picasso Cafe: A case study evaluation of two work-based training programs for disadvantaged youth, Social Research and Demonstration Corporation, Ottawa, Ontario.

Curtis, L.J. (2001) Lone motherhood and health status, Canadian Public Policy 27(3): 335–356.

Danziger, S. (1996) The real war on crime: The report of the National Criminal Justice Commission, Harper Collins, New York.

Danziger, S.K., M.J. Carlson, and J.R. Henly (2001) Post-welfare employment status is a significant factor for psychological well-being, in M.C. Lennon (ed), Welfare, work and well-being, Haworth Medical Press, New York.

Danziger, S., M. Corcoran, S. Danziger, C. Heflin, A. Kalil, J. Levine, D. Rosen, K. Seefeldt, K. Siefert, and R. Tolman (2000) Barriers to the employment of welfare recipients, in R. Cherry and W. M. Rodgers, III (eds), Prosperity for all? The economic boom and African Americans, Russell Sage Foundation, New York.

Danziger, S.K., A. Kalil, and N.J. Anderson (2000) Human capital, physical health, and mental health of welfare recipients: Co-occurrence and correlates, Journal of Social Issues, 56(4): 635–654.

Danziger, S. and K.S. Seefeldt (2000) Ending welfare through work first: Manager and client views, Families in Society: The Journal of Contemporary Human Services 81(6): 593–604.

Davison, H.K. and M.J. Burke (2000) Sex discrimination in simulated employment contexts: A meta-analytic investigation, Journal of Vocational Behavior 56(2): 225–248.

Dawson, T., S. Dickens, and S. Finer (2000) New Deal for lone parents: Report on qualitative interviews with individuals, Research and Development Report ESR55, Employment Service, Sheffield.

DeBord, K., R. Fanning Canu, and J.L. Kerpelman (2000) Understanding a work-family fit for single parents: Moving from welfare to work, Social Work 45(4): 313–324.

Department of Labour and Ministry of Social Development (2001) Evaluating the February 1999 Domestic Purposes Benefit and Widows Benefit Reforms, Wellington.

Department for Work and Pensions (UK) (2003) Client group analysis, Quarterly bulletin on the population of working age on key benefits – August 2002, Information Centre, Analytical Services Directorate, Newcastle upon Tyne.

De Radd, J.P. (1997) The duration and repeat of unemployment benefit spells: A study of a cohort of New Zealand unemployment beneficiary spells 1992–1995, unpublished Master’s thesis, Victoria University, Wellington.

Derr, M.K., S. Douglas, and L. Pavetti (2001) Providing mental health services to TANF recipients: Program design choices and implementation challenges in four states, paper presented at the Joint Center for Research on Poverty conference The Hard-to-Employ and Welfare Reform, 28 February – 1 March 2002, Washington, DC.

Dockery, A.M. and E. Webster (2001) Long-term unemployment and work deprived individuals: Issues and policies, Centre for Economic Policy Research Discussion Paper No. 445, Australian National University, Canberra.

Dooley, D. and J. Prause (1998) Underemployment and alcohol misuse in the National Longitudinal Survey of Youth, Journal of Studies on Alcohol 59(6): 669–680.

Dorset, R. (2001) Workless couples: Characteristics and labour market transitions, Employment Services Report ESR 79, Employment Service, Sheffield.

Dorset, R., L. Finlayson, R. Ford, A. Marsh, M. White and G. Zarb.(1998) Leaving incapacity benefit, Research Report No 86, Department of Work and Pensions, Corporate Document Services, Leeds.

Drake, R.E. and M.A. Wallach (2000) Dual diagnosis: 15 years of progress, Psychiatric Services 51(9): 1126–1129.

Duncan, P., E. Kerekere and D. Malaulau (1996) Women on low incomes. The Young Women’s Christian Association of Aotearoa-New Zealand and the New Zealand Council of Christian Social Services, Wellington.

Edin, K. and L. Lein (1997) Making ends meet: How single mothers survive welfare and low-wage work, Russell Sage Foundation, New York.

Eissa, N. and J.B. Liebman (1996) Labor supply response to the Earned Income Tax Credit, Quarterly Journal of Economics, 111(2): 605–37.

Ellwood, D. (2000) The impact of the Earned Income Tax Credit and social policy reforms on work, marriage, and living arrangements, Joint Center for Poverty Research Working Paper No. 124, Northwestern University / University of Chicago, Chicago, Illinois.

Encel, S. (2000) Mature age unemployment: A long-term cost to society, Economic and Labour Relations Review 11(2): 233–45.

Evans, H. (2000) Sprouting seeds: Outcomes from a community-based employment programme, Centre for Analysis of Social Exclusion Report No. 7, London School of Economics, London.

Fagnoni, C.M. (2001) Moving hard to employ recipients into the workforce, GAO-01-368, FDCH Government Accounting Report, Washington, DC.

Farkas, G. and P. England (1994) Industries, firms, and jobs: Sociological and economic approaches (expanded edn), Aldine de Gruyter, New York.

Fergusson, D.M., L.J. Horwood and M.T. Lynskey (1997) The effects of unemployment on psychiatric illness during young adulthood, Psychological Medicine 27(2): 371–381.

Finch, H. and W. O’Connor, with J. Millar, J. Hales, A. Shaw, and W. Roth (1999) New Deal for lone parents: Learning from the prototype areas, Research Report No 92, Department of Work and Pensions, Corporate Document Services, Leeds,

Finlayson, L., R. Ford, A. Marsh, S. McKay, and A. Mukherjee (2000) The British lone parent cohort 1991–1998, Research Report No. 128, Department of Work and Pensions, Corporate Document Services, Leeds.

Finlayson, L. and A. Marsh (1998) Lone parents on the margins of work, in R. Ford, and J. Millar (eds), Private lives and public responses: Lone parenthood and future policy in the UK Policy Studies Institute, London, pp.193–207.

Fletcher, C.N., S. Garashky and H.H. Jensen (2002) Transitioning from welfare to work: No bus, no car, no way, paper presented at the Joint Center for Poverty Research conference The Hard-to-Employ and Welfare Reform, 28 February to 1 March, Washington, DC.

Ford, R. (1996) Childcare in the balance. How lone parents make decisions about work, Policy Studies Institute, London.

Ford, R., A. Marsh and L. Finlayson (1998) What happens to lone parents, Research Report No. 77, Department of Work and Pensions, Corporate Document Services, Leeds.

Foreman, A., M. Hawtin, and K. Ward (1997) A springboard away from dependency, Social Policy Journal of New Zealand 8: 155–165.

Fraker, T. and C. Prindle (1996) Findings from the evaluation of Iowa’s Limited Benefit Plan, paper presented at the Annual Research Conference of the Association for Public Policy and Management, Pittsburgh, PA.

Friedlander, D. and G. Burtless (1995) Five years after: The long-term effects of welfare-to-work programs, Russell Sage Foundation, New York.

Gallie D. (1994) Are the unemployed an underclass? Some evidence from the Social Change and Economic Life Initiative, Sociology 28(3): 737–757.

Gardiner, K. (1997) Bridges from benefit to work: A review, Joseph Rowntree Foundation, York.

Gardiner, P., M. Fletcher, P. Mersi, and A. Reynolds (1994) An analysis of the dynamics of unemployed exit probabilities and repeat spells, unpublished paper, Labour Market Analysis Unit, Department of Labour, Wellington.

Glennerster, H., R. Lupton, P. Noden, and A. Power (1999) Poverty, social exclusion and neighbourhood: Studying the area bases of social exclusion, Centre for Analysis of Social Exclusion Paper No. 22, London School of Economics, London.

Gobbi, M. and D. Rea (2002) The unemployment spells of registered job seekers, Labour Market Bulletin, 2000–2002 Special Issue: 31–56.

Gottschalk, P. and R.A. Moffitt (1994) Welfare dependence: Concepts, measures, and trends, American Economic Association Papers and Proceedings 84(2): 39–42.

Graham, P., A. Jordon, and B. Lamb (1990) An equal chance? Or no chance? A study of discrimination against disabled people in the labour market, The Spastics Society, London.

Greenwood, J. and J-P. Voyer (2000) Experimental evidence on the use of earnings supplements as a strategy to ‘make work pay’, OECD Economic Studies 31: 43–67.

Gregg, P., P. Johnson and H. Reed (1999) Entering work and the British tax and benefit system, The Institute for Fiscal Studies, London.

Grogger, J. (1995) The effect of arrest on employment and earnings of young men, Quarterly Journal of Economics 110: 51–71.

Hales, J., W. Roth, M. Barnes, J. Millar, C. Lessof, M. Gloyer, and A. Shaw (2000) Evaluation of the New Deal for Lone Parents: Early Lessons from the Phase One Prototype – Findings of Surveys, Research Report No 109, Department of Work and Pensions, Corporate Document Services, Leeds.

Hall, W., M. Teesson, M. Lynskey, and L. Degenhardt (1999) The 12-month prevalence of substance use and ICD-10 substance use disorders in Australian adults: Findings from the National Survey of Mental Health and Well-Being, Addiction 94(10): 1541–1550.

Hand, K. (2001) Welfare to Work Strategies: A review of evidence of the effectiveness of selected countries’ policies and programmes, report prepared for the Ministry of Social Development, Wellington.

Hand, K. (2002) Financial employment disincentives facing working age welfare recipients in New Zealand, unpublished background paper, Ministry of Social Development, Wellington.

Hannan, C. (1998) Beyond networks: ‘Social cohesion’ and unemployment exit rates, unpublished paper, Institute for Labour Research, University of Essex.

Harris, K.M. (1993) Work and welfare among single mothers in poverty, American Journal of Sociology 99: 317–352.

Harris, K.M. (1996) Life after welfare: Women, work, and repeat dependency, American Sociological Review 61: 407–426.

Hedges, A. and W. Sykes (2001) Moving between sickness and work, Research Report No 151, Department of Work and Pensions, Corporate Document Services, Leeds.

Hershey, A. M. and L. Pavetti (1997) Turning job finders into job keepers, The Future of Children, 7(1): 74–86.

Heymann, S.J. and A. Earle (1997) Working conditions faced by poor families and the care of children, Focus 19(1): 56–58, Institute for Research on Poverty, University of Wisconsin, Madison, WI.

Hirsch, D. (1999) Welfare beyond work: Active participation in a new welfare state, Joseph Rowntree Foundation, York.

Holterman, S., J. Brannen, P. Moss, and C. Owen (1999) Lone parents and the labour market: Results from the 1997 Labour Force Survey and review of research, Research Report No 23, Department for Work and Pensions, Research and Analysis Publications, Sheffield.

Holzer, Harry J. (1996) What Employers Want: Job Prospects for Less-Educated Workers. New York: Russell Sage Foundation.

Holzer, H.J. (2000) Unemployment insurance and welfare recipients: What happens when the recession comes? No A-46 in series New Federalism: Issues and Options for the States, The Urban Institute, Washington, DC.

Holzer, H.J. and S. Danziger (1998) Are jobs available for disadvantaged workers in urban areas? Institute for Research on Poverty Discussion Paper No 1157-98, University of Wisconsin, Madison, WI.

Holzer, H.J. and R. J. LaLonde (1999) Job change and job stability among less-skilled young workers, Institute for Research on Poverty Discussion Paper No 1191‑99, University of Wisconsin, Madison, WI.

Holzer, H.J and M.A. Stoll (2000) Employer demand for welfare recipients by race, Assessing the New Federalism Discussion Paper No 01-07, The Urban Institute, Washington, DC.

Holzer H.J, M.A. Stoll, and D. Wissoker (2001) Job performance and retention among welfare recipients, Joint Centre for Poverty Research Working Paper No 231, Northwestern University / University of Chicago, Chicago, Illinois.

Hoynes, H.W. and R. Moffitt (1995) The effectiveness of financial work incentives in disability insurance and Social Security Insurance: Lessons from other transfer programs, Institute for Research on Poverty Discussion Paper #1073-95, University of Wisconsin, Madison, WI.

Hunter, B. (1996) Explaining changes in the social structure of employment: The importance of geography, Social Policy Research Centre Discussion Paper No 67, University of New South Wales, Sydney.

Iacovou, M. and R. Berthoud (2000) Parents and employment, Research Report No 107, Department of Work and Pensions, Corporate Document Services, Leeds.

Institute for Research on Poverty (1997) Children with disabilities and the SSI program, Focus 19(1): 51.

Jackson, A.P., M. Tienda and C. Huang (2001) Capabilities and employability of unwed mothers, Children and Youth Services Review 23(4/5): 327–351.

Janlert, U. (1997) Unemployment as a disease and diseases of the unemployed, Scandinavian Journal of Work Environment and Health 23: 79–83.

Jargowsky, P.A. (1997) Poverty and place: Ghettos, barrios, and the American city, Russell Sage Foundation, New York.

Johnson, C. and S. Fielding (2000) First effects of ONE: Qualitative research with clients – Stage One, Research Report No 126, Department of Work and Pensions, Corporate Document Services, Leeds.

Johnson, R. (2001) Welfare recipients’ road to economic self-sufficiency: A study of the job transition patterns of welfare recipients, unpublished paper, University of Michigan, Ann Arbor, MI.

Jonczyk, M. and T. Smith (1990) Crossing the barriers: A survey of the barriers to employment for DSS clients with disabilities, Policy Research Paper No 58, Department of Social Security, Social Policy Division, Canberra.

Jordan, A. (1989) Lone parents and wage earners? Employment prospects of sole parent pensioners, Background paper No 31, Social Security Review, Department of Social Security, Canberra.

Joshi, H., S. Macran, and S. Dex (1996) Employment after childbearing and women’s subsequent labour force participation: Evidence from the British 1958 birth cohort, Journal of Population Economics 9(3): 325–348.

Kalil, A., M.E. Corcoran, S.K. Danziger, R. Tolman, K.S. Seefeldt, D. Rosen, and Y. Nam (1998) Getting jobs, keeping jobs, and earning a living wage: Can welfare reform work? Institute for Research on Poverty Discussion Paper No 1170-98, University of Wisconsin, Madison, WI.

Kapsalis, C. (1998) The connection between literacy and work: Implications for social assistance recipients, Working Paper W-98-1E, Applied Research Branch, Strategic Policy, Human Resources Development Canada, Quebec.

Kassarda, J.D. and K.F. Ting (1996) Joblessness and poverty in America’s central cities: Causes and policy prescriptions, Housing Policy Debate 7(2): 387–419.

Kempson, E. (1996) Life on a Low Income, Joseph Rowntree Foundation, York.

Kempson, E., J. Ford, and J. England (1997) In-work benefits: Do they assist the transition into work? Benefits (Articles) 19 (April/May): 20–23.

Kennelly, I. L. (1995) That could come from a lack of nurturing: Employers’ images of African-American women, unpublished Master’s thesis, Department of Sociology, University of Georgia, Atlanta.

Khan, S., R.P. Murray, and G.E. Barnes (2002) A structural equation model of the effect of poverty and unemployment on alcohol abuse, Addictive Behaviors 27(3): 405–423.

Kimmel, J. (1995) The effectiveness of child-care subsidies in encouraging the welfare-to-work transition of low-income single mothers, American Economic Association Papers and Proceedings 85(2): 271–275.

Kirshenman, J. and K.M. Neckerman (1991) ‘We’d love to hire them but …’: The meaning of race for employers, in C. Jencks and P. Peterson (eds), The Urban Underclass, Brookings Institution, Washington, DC.

Klepinger, D., S. Lundberg, and R. Plotnick (1997) How does adolescent fertility affect the human capital and wages of young women? Institute for Research on Poverty Discussion Paper No 1145-97, University of Wisconsin, Madison, Wisconsin.

Kling, J.R., J.B. Liebman, and L.F. Katz (2001) Bullets don’t got no name: Consequences of fear in the ghetto, Joint Centre for Poverty Research Working Paper 225, Northwestern University / University of Chicago, Chicago, Illinois.

Laudet, A.B., S. Magura, H.S. Vogel, and E.L. Knight (2002) Interest in and obstacles to pursuing work among unemployed dually diagnosed individuals, Substance Use and Misuse 37(2): 145–170.

Leahy, P.J., T.F. Buss, and J.M. Quane (1995) Time on welfare: Why do people enter and leave the system? American Journal of Economics and Sociology 54(1): 33–46.

Leete, L., N. Bania and C. Coulton (1999) Distance and commute times to work for welfare exiters, Center on Urban Poverty and Social Change Briefing Report No 9908, Case Western Reserve University, Cleveland, OH.

Lennon, M.C. (ed) (2001) Welfare, work and well-being, Haworth Medical Press, New York.

Levesque, M. and D. White (2001) Social capital, human capital and exit of beneficiaries from long-term social assistance, Canadian Journal of Sociology 26(2): 167–192.

Levine, M., H. Wyn and L. Asiasiga (1993) Lone parents and paid work: A study of employment patterns and barriers, and options for change, Social Policy Agency, Wellington.

Lewis, J., L. Mitchell, T. Sanderson, W. O’Connor, and M. Clayden (2000) Lone parents and personal advisers: Roles and relationships, Research Report No 122, Department of Work and Pensions, Corporate Document Services, Leeds.

Lichter, D.T. and R. Jayakody (2002) Welfare reform: How do we measure success? Annual Review of Sociology 28: 117–141.

Lloyd, S. and N. Taluc (1999) The effects of male violence on women’s employment, Violence Against Women 5: 370–392.

Loprest, P. and G.Acs (1995) Profile of disability among families on AFDC, Urban Institute, Washington DC.

Loprest, P. and S.R. Zedlewski (2002) Making TANF work for the hard to serve, Short Takes on Welfare Policy Number 2, Assessing the New Federalism series, The Urban Institute, Washington, DC.

Loumidis, J., R. Youngs, C. Lessof, and B. Stafford (2001) New Deal for disabled people: National Survey of Incapacity Benefits Claimants, Research Report No 160, Department of Work and Pensions, Corporate Document Services, Leeds.

Lunt, N. and R. Pernice (1999) New Zealand disability employment policy in the 1990s, Social Policy Journal of New Zealand, 12: 16–34.

Maloney, T. (1997) Benefit reforms and labour market behaviour in New Zealand, Institute of Policy Studies, Victoria University of Wellington.

Marcenko, M.O. and J. Fagan (1996) Welfare to work: What are the obstacles? Journal of Sociology and Social Welfare 23(3): 113–131.

Marsh, A. (1997) Lowering the barriers to work in Britain, Social Policy Journal of New Zealand 8: 111-135

Marsh, A., S. McKay, A. Smith, and A. Stephenson (2001) Low-income families in Britain: Work, welfare and social security in 1999, Research Report No 138, Department of Work and Pensions, Corporate Document Services, Leeds.

Massey, D. (2001) The prodigal paradigm returns: Ecology comes back to sociology, in A. Booth and A.C. Crouter (eds), Does It Take a Village? Community Effects on Children, Adolescents, and Families, Lawrence Erlbaum Associates, Mahwah, NJ, pp. 41–48.

McHugh, M. and J. Millar (1996) Sole mothers in Australia: Supporting mothers to seek work, Social Policy Research Centre Discussion Paper No. 71, University of New South Wales, Sydney.

McKay, S. (2002) Low/moderate-income families in Britain: Work, Working Families’ Tax Credit, and childcare in 2000, Research Report No. 161, Department of Work and Pensions, Corporate Document Services, Leeds.

McKay, A. and S. Marsh (1994) Lone parents and work, Research Report No. 25, Department of Work and Pensions, Corporate Document Services, Leeds.

McKay, S., A. Smith, R. Youngs, and R. Walker (1999) Unemployment and job seeking after the introduction of Job seeker’s Allowance, Research Report No. 99, Department of Work and Pensions, Corporate Document Services, Leeds.

McKay, S., R. Walker, and R. Youngs (1997) Unemployment and job seeking before Job seeker’s Allowance, Research Report No. 73, Department of Work and Pensions, Corporate Document Services, Leeds.

McLaughlin, D.K. and D.T. Lichter (1997) Poverty and the marital behavior of young women, Journal of Marriage and the Family 59(3): 582–94.

Mead, L. (1986) Beyond entitlement: The social obligations of citizenship, Free Press, New York.

Meyer, B.D. and D.T. Rosenbaum (2001) Welfare, the Earned Income Tax Credit, and the labor supply of single mothers, Quarterly Journal of Economics 116(3): 1063–1114.

Meyer, B.D. and D.T. Rosenbaum (2000) Making single mothers work: Recent tax and welfare policy and its effects, National Tax Journal 53(4): 1027–61.

Meyers, M.K. (1993) Child care in JOBS employment and training program: What difference does quality make? Journal of Marriage and the Family 55: 767–783.

Meyers, M., H. Brady and E.Y. Seto (2002) Expensive children in poor families: The intersection of childhood disabilities and welfare, paper presented at the Joint Center for Research on Poverty conference The Hard-to-Employ and Welfare Reform, 28 February – 1 March, Washington, DC.

Michalopoulos, C. and G. Berlin (2000) Financial work incentives for low-wage workers, Joint Center for Poverty Research Policy Brief Vol. 3, No. 4, Northwestern University / University of Chicago, Chicago, Illinois.

Michalopoulos, C et al (2002) Making Work Pay: Final Report on the Self-Sufficiency Project for Long-Term Welfare Recipients. Ottawa: Social Research and Demonstration Corporation.

Millar, J. (1994) Understanding labour supply in context: households and incomes, in A. Bryson and S. McKay (eds), Is it Worth Working? Policy Studies Institute, London.

Millar, J. and T. Ridge (2001) Families, poverty, work, and care: A review of the literature on lone parents and low-income couple families with children, Research Report No. 153, Department of Work and Pensions, Corporate Document Services, Leeds.

Mills, B.F., J.R. Alwang, and G. Hazarika (2001) Welfare reform and the well-being of single female-headed families: A semi-parametric analysis, Review of income and Wealth 47(1): 81–104.

Moffitt, R.A. (2001) Experienced-based measures of heterogeneity in the welfare caseload, in M. Ver Ploeg, R.A. Moffitt, and C.F. Citro (eds), Studies in welfare populations: Data collection and research issues, National Academy Press, Washington.

Moffitt, R.A. and D. Stevens (2001) Changing caseloads: Macro influences and micro composition, revised version of paper presented at the conference Welfare Reform Four Years Later: Progress and Prospects, Federal Reserve Bank of New York, 17 November, 2000.

Montgomery, S.M., D.G. Cook, M.J. Bartley, and M.E.J. Wadsworth (1998) Unemployment, cigarette smoking, alcohol consumption and body weight in young British men, European Journal of Public Health, 8(1): 21–27.

Morrison, P.S. (1993) Exploring contemporary gender and ethnicity issues: Making (much better) use of Supermap2, Working Paper No 22, Victoria University, Wellington.

Moss, P. and C. Tilly (2001) Soft skills and race, working paper, Russell Sage Foundation, New York.

Mullahy, J. and J. Sindelar (1996) Employment, unemployment, and problem drinking, Journal of Health Economics 15(4): 408–434.

Noble, M., G. Smith, S.Y. Cheung (1998) Lone mothers moving in and out of benefits, York Publishing Services for the Joseph Rowntree Foundation, York.

Olson, K. and L. Pavetti (1996) Personal and family challenges to the successful transition from welfare to work, The Urban Institute, Washington, DC.

Orthner, D. K. and P.A. Neenan (1996) Children’s impact on stress and employability of mothers in poverty, Journal of Family Issues 17(5): 667–687.

Parker, L. (1994) The role of workplace support in facilitating self-sufficiency among single mothers on welfare, Family Relations 43: 168–173.

Parker, B. (1997) Very long-term job seekers’ barriers to employment: A nationwide survey, Labour Market Bulletin 1: 63–79.

Patel, N. (2001) Workforce development: Employment retention and advancement under TANF, Center for Law and Social Policy Publication No. 01-27, Washington, DC.

Patterson, J. (1999) The third age – the age of reason: A gift and not a burden, Social Policy Journal of New Zealand 13: 1–13.

Pavetti, L. (1997) Against the odds: Steady employment among low-skilled women, The Urban Institute, Washington, DC.

Pavetti, L. (1999) How much more can welfare mothers work? Focus 20(2): 16–19, Institute for Research on Poverty, University of Wisconsin, Madison, WI.

Pavetti, L., K. Olson, D. Nightingale, A. Duke and J. Isaacs (1997) Welfare-to-work options for families facing personal and family challenges: Rationale and program strategies, The Urban Institute, Washington, DC.

Pernice R. and N. Long (1996) Long-term unemployment, employment attitudes and mental health, Australian Journal of Social Issues 31(3): 311–326.

Petersen, T., I. Saporta, and M.D.L. Seidel (2000) Offering a job: Meritocracy and social networks, American Journal of Sociology 106(3): 763–816.

Petterson, S.M. and L. Friel (2001) Psychological distress and hopelessness of women on welfare related to material hardship, in M.C. Lennon (ed), Welfare, work and well-being, Haworth Medical Press, New York.

Polit, D., A. London, and J. Martinez (2001) The health of poor urban women: Findings from the Project on Devolution and Urban Change, Manpower Demonstration Research Corporation, New York.

Polit, D.F., R. Widom, K. Edin, S. Bowie, A.S. London, E.K. Scott, and A. Valenzuela (2001) Is work enough? The experiences of current and former welfare mothers who work, Manpower Demonstration Research Corporation, New York.

Pollack, H.A., S. Danziger, K.S. Seefeldt, and R. Jayakody (2002) Substance use among welfare recipients: trends and policy responses, paper presented at the Joint Center for Poverty Research conference The Hard-to-Employ and Welfare Reform, 28 February to 1 March, Washington, DC.

Puniard, A. and C. Harrington (1993) Working through the poverty traps: Results of a survey of sole parent pensioners and unemployment beneficiaries, Social Security Journal December: 1–17.

Quint, J., J. Musick, and J. Ladner (1994) Lives of promise, lives of pain: Young mothers after New Chance, Manpower Demonstration Research Corporation, New York.

Rangarajan, A., A. Meckstroth, and T. Novak (1998) The effectiveness of the Postemployment Services Demonstration: Preliminary findings, Mathematica, Princeton, NJ.

Rangarajan, A., P. Schochet, and D. Chu (1998) Employment experiences of welfare recipients who find jobs: Is targeting possible? Mathematica, Princeton, NJ.

Rank, M.S. (1994) A view from the inside out: Recipients’ perception of welfare, Journal of Sociology and Social Welfare 21: 27–47.

Raphael, J. (1996) Domestic violence: Telling the untold welfare-to-work story, Taylor Institute, Chicago.

Raphael, S., M.A. Stoll, and H.J. Holzer (2000) Are suburban firms more likely to discriminate against African-Americans? Journal of Urban Economics 48(3): 485–508.

Ravaud, J., B. Madiot, and I. Ville (1992) Discrimination towards disabled people seeking employment, Social Services and Medicine 35(8): 951–8.

Regenstein, M., J.A. Meyer, and J.D. Hicks (1998) Job prospects for welfare recipients: Employers speak out, New Federalism Series A., No. A-25, The Urban Institute, Washington, DC.

Reskin, B.F. and P.A. Roos (1990) Job queues, gender queues: Explaining women’s inroads into male occupations, Temple University Press, Philadelphia, PA.

Riccio, J and S. Freedman (1995) Can they all work? A study of the employment potential of welfare recipients in a welfare-to-work program, Working paper, Manpower Demonstration Research Corporation, New York.

Riccio, J., D. Friedlander, and S. Freedman (1994) GAIN: Benefits, costs and three-year impacts of a welfare-to-work program, Manpower Demonstration Research Corporation, New York.

Rodriguez, E., E.A. Frongillo, and P. Chandra (2001) Do social programmes contribute to mental well-being? The long-term impact of unemployment on depression in the United States, International Journal of Epidemiology 30(1): 163–170.

Rodriguez, E., K. Lasch, and J.P. Mead (1997) The potential role of unemployment benefits in shaping the mental health impact of unemployment, International Journal of Health Services 27(4): 601–623.

Rosenheck, R. (1997) Disability payments and chemical dependence: Conflicting values and uncertain effects, Psychiatric Services 48(6): 789–791.

Sandefur, G.D. and S.T. Cook (1998) Permanent exits from public assistance: The impact of duration, family, and work, Social Forces 77(2): 763–786.

Sansone, F. (1998) Social supports’ contribution to reduce welfare dependency: Program outcomes of long term welfare recipients, Journal of Sociology and Social Welfare 25(4): 105–126.

Schumacher, R. and M. Greenberg (1999) Child care after leaving welfare: early evidence from state studies, Centre for Law and Social Policy, Washington DC.

Scott, E.K., K. Edin, A.S. London, and J.M. Mazelis (2001) My children come first: Welfare-reliant women’s post-TANF views of work-family trade-offs and marriage, in G.J. Duncan and P.L. Chase-Landsdale (eds), For Better and For Worse: Welfare Reform and the well being of children and families, Russell Sage Foundation, New York.

Scott, E.K., A.S. London and K. Edin (2000) Looking to the future: Welfare-reliant women talk about their job aspirations in the context of welfare reform, Journal of Social Issues 56(1): 727–746.

Scott, G., A. McKay, C. Sawers, and R. Harris (1999) What can we afford? A woman’s role. Considering the welfare implications of money, Research Findings No. 1, Scottish Poverty Information Unit, Glasgow Caledonian University, Glasgow.

Scottish Poverty Information Unit (1998) Disability and poverty, Briefing No. 7, Glasgow Caledonian University, Glasgow.

Seguino, S. and S.S. Butler (1998) To work or not to work: Is that the right question? Review of Social Economy 66(2): 190–219.

Seth-Purdie, R. (2000) Multiple risk exposure and likelihood of welfare receipt: Social policy and human capital, Family Matters 57: 46–53.

Shaver, S., A. King, M. McHugh, and T. Payne (1994) At the end of eligibility: Female sole parents whose youngest child turns 16, Social Policy Research Centre Reports and Proceedings No. 117, University of New South Wales, Sydney.

Shaw, A. R. Walker, K. Ashworth, S. Jenkins, and S. Middleton (1996) Moving off income support: Barriers and bridges, Research Report No. 53, Department of Work and Pensions, Corporate Document Services, Leeds.

Sheehan, M. and M. Tomlinson (1998) Unemployment duration in an unemployment blackspot, Labour 12(4): 643–673.

Singley, S.G. (1995) Gender Inequalities in the Labor Market and Poverty among Working Single Mothers, unpublished Master’s Thesis, Department of Sociology, The Pennsylvania State University, University Park, Pennsylvania.

Skogman, P. (1998) Occupational attainment and earnings for disabled workers, Working Paper 3/1998, Institutet för Social Forskning, Stockholms Universitet.

Snape, D. (1998) Recruiting long-term unemployed people, Research Report No. 76, Department for Work and Pensions, Corporate Document Services, Leeds.

Social Exclusion Unit (2003) Making the connections: Transport and social exclusion, Office of the Deputy Prime Minister, London.

Soldera, P. (1999) Spatial hysteresis and the clustering of urban unemployment, Master’s thesis, Department of Geography, Victoria University, Wellington.

Sparkes, J. (1999) Schools, education, and social exclusion, Centre for Analysis of Social Exclusion Paper 29, London School of Economics, London.

Stafford, B., C. Heaver, N. Croden, A.A. Smith, S. Maguire and J. Vincent, Research Report No 79, Department for Work and Pensions, Corporate Document Services, Leeds.

Statistics New Zealand (2002) Labour Market Statistics, 2001 – Reference Reports, Wellington.

Stewart, J.M., R.W. McQuaid, A.C. Charlesworth-May, and J.Adams (1998) A summary of the results of a survey of unemployed workers in the Auckland Eastern Suburbs TTWA, Labour Market Bulletin 1&2: 145–161.

Stewart, L.D. and R. Perlow (2001) Applicant race, job status, and racial attitudes as predictors of employment discrimination, Journal of Business Psychology 16(2): 259–275.

Stawn, J., M. Greenberg and S. Savner (2001) Improving employment outcomes under TANF, paper prepared for the Conference on the New World of Welfare, Washington DC, 1–2 February 2001.

Strawn, J. and K. Martinson (2000) Steady work and better jobs: How to help low-income parents sustain employment and advance in the workforce, Manpower Demonstration Research Corporation, New York.

Stephenson, A. (2001) Work and welfare: Attitudes, experiences, and behaviour of nineteen low-income families, In-house Report No. 76, Social Research Division, Department for Work and Pensions, London.

Swartz, J.A., A.J. Lurigio and P. Goldstein (2000) Severe mental illness and substance use disorders among former Supplemental Security Income beneficiaries for drug addiction and alcoholism, Archives of General Psychiatry 57(7): 701–707.

Sweeney, E.P. (2000) Recent studies indicate that many parents who are current or former welfare recipients have disabilities or other medical conditions, Center on Budget and Policy Priorities, Washington, DC.

Terza, J.V. (2002) Alcohol abuse and employment: A second look, Journal of Applied Econometrics 17(4): 393–404.

Thomas, A., N. Pettigrew, D. Cotton, and P. Tovey (1999) Keeping in touch with the labour market, Research Report No. 96, Department of Work and Pensions, Corporate Document Services, Leeds.

Tolman, R.M., S.K. Danziger, and D. Rosen (2002) Domestic violence and economic well-being of current and former welfare recipients, paper presented at the Joint Center for Research on Poverty conference The Hard-to-Employ and Welfare Reform, 28 February – 1 March, Washington, DC.

Trickey, H., K. Kellard, R. Walker, K. Ashworth, and A. Smith (1998) Unemployment and job seeking: Two years on, Research Report No. 87, Department of Work and Pensions, Corporate Document Services, Leeds.

US Department Health and Human Services and US Department of Education (2001a) How effective are different welfare-to-work approaches? Five-year adult and child impacts for eleven programs, National Evaluation of Welfare-to-Work Strategies series, Washington, DC.

US Department Health and Human Services and US Department of Education (2001b) What works best for whom? Impacts of 20 welfare-to-work programs by subgroup, National Evaluation of Welfare-to-Work Strategies series, Washington, DC.

US General Accounting Office (1999) Welfare reform: Assessing the effectiveness of various welfare-to-work approaches, Report to the Congressional Committees, GAO/HEHS-99-179, Washington, DC.

US General Accounting Office (2001) Welfare reform: Moving hard-to-employ recipients into the workforce, Report to the Chairman, Subcommittee on Human Resources, Committee on Ways and Means, House of Representatives, GAO-01-368, Washington DC.

Vandenheuvel, A. (1996) The relationship between women’s working arrangements and their child care arrangements, Australian Bulletin of Labour 22(4): 288–305.

Vartanian, T.P. and J.M. McNamara (2000) Work and economic outcomes after welfare, Journal of Sociology and Social Welfare 27(2): 41–77.

Wadsworth, M.E.J., M.J. Bartley, and S.M. Montgomery (1999) The persisting effect of unemployment on health and social wellbeing in men early in working life, Social Science and Medicine 48(10): 9.

Waldinger, R.D. (2003) How the other half works: Immigration and the social organization of labor, University of California Press, Berkeley.

Walker, A. and P. Taylor (1998) Combating age barriers in employment: A European portfolio of good practice, European Foundation for the Improvement of Living and Working Conditions, Office for Official Publications of the European Communities, Luxembourg.

Walker, R. and K. Kellard (2001) Staying in work: A new policy agenda? In K. Kellard (ed), Lone Parents and Sustainable Employment: Proceedings from a one-day seminar, 1 May, Centre for Research on Social Policy, Loughborough University, Loughborough, Leicestershire.

Watson, R., D. Mare and P. Gardiner (1997) Predicting the duration of unemployment spells, Labour Market Bulletin 2: 51–65.

Webster, D. (2000) The geographical concentration of labour-market disadvantage, Oxford Review of Economic Policy 16(1): 114–28.

Western, B. (2001) Incarceration, unemployment, and inequality, Focus 21(3): 32–36, Institute for Research on Poverty, University of Wisconsin, Madison, WI.

White, M. and J. Forth (1998) Pathways through unemployment: The effects of a flexible labour market, York Publishing Services for the Joseph Rowntree Foundation, York.

Williams, C.C. and J. Windebank (2000) Helping people to help themselves: Policy lessons from a study of deprived urban neighbourhoods in Southampton, Journal of Social Policy 29(3): 355–373.

Wilson, M. (1996) Income Support for the Unemployed and Labour Supply Behaviour: Three Frameworks for Analysis, Social Policy Agency, Wellington.

Wilson, M. (2000) The policy response to the Employment Task Force and changing patterns of Domestic Purposes Benefit receipt: A cohort analysis, Social Policy Journal of New Zealand 14: 78–103.

Wilson, M. (2001) Longitudinal analysis of administrative data: The Ministry of Social Policy’s Benefits Dynamics Project, paper presented at the Ministry of Social Policy Longitudinal Research and Social Policy Seminar, Wellington, 5–6 April.

Wilson, M. and D. Ball (2000) Policy change and patterns of Domestic Purposes Benefit receipt: A multiple cohort analysis using Benefit Dynamics data, paper presented at the Labour, Employment and Work Conference, Victoria University of Wellington, 23 November.

Wilson, W.J. (1987) The truly disadvantaged: The inner city, the underclass, and public policy, The University of Chicago Press, Chicago.

Wilson, W.J. (1996) When Work Disappears: The World of the New Urban Poor, Knopf, New York.

Wilson, W.J. (1998) When work disappears: New implications for race and urban poverty in the global economy, Centre for Analysis of Social Exclusion Paper No. 17, London School of Economics, London.

Woodfield, K. and H. Finch (1999) New Deal for Lone Parents: Evaluation of innovative schemes, Research Report No. 89, Department of Work and Pensions, Corporate Document Service, Leeds.

Wylie, C.R (1980) Factors affecting the participation in the workforce of female heads of one parent families, Department of Social Welfare, Wellington.

Zanetti, C. (1994) Sole parents: Trends and issues, Social Security Journal December, 92–102.

Zedlewski, S.R. (1999) Work-related activities and limitations of current welfare recipients, Discussion Paper No. 99-06, Assessing the New Federalism series, The Urban Institute, Washington, DC.

Zedlewski, S.R. and D.W. Alderson (2001) Before and after reform: How have families on welfare changed? Assessing the New Federalism, Series B, No. B-32, The Urban Institute, Washington, DC.

6. Appendix

Below is a list of the policy research institutes and government departments consulted, and their websites. All relevant materials readily available from these sources between January and April 2002 were reviewed. As needed, other literature cited in these original sources was also consulted, as was the general academic literature for major review articles on related topics, and for filling in gaps left by the first-level review.

United States

Brookings Institution – policy briefs www.brook.edu

Center on Budget and Policy Priorities www.cbpp.org

Center for Law and Social Policy www.clasp.org

Center on Urban Poverty and Social Change, Case Western Reserve University http://povertycenter.cwru.edu

Child Trends www.childtrends.org

Economic Policy Institute www.epinet.org

Institute for Research on Poverty, U Wisconsin www.ssc.wisc.edu/irp

Institute for Women’s Policy Research www.iwpr.org

Joint Center for Poverty Research, Northwestern University and University of Chicago www.jcpr.org

Manpower Demonstration Research Corporation www.mdrc.org

Mathematica Policy Research, Inc. www.mathematica-mpr.com

Michigan Program on Poverty and Social Welfare Policy www.ssw.umich.edu/poverty

Moving Ideas: The Electronic Policy Network www.movingideas.org

National Center for Children in Poverty, Columbia University www.nccp.org

Office of the Assistant Secretary for Planning and Evaluation, Department of Health and Human Services http://aspe.hhs.gov/hsp

Radcliffe Public Policy Center www.radcliffe.edu/pubpol/index.html

RAND www.rand.org

Research Forum on Children, Families, and the New Federalism www.researchforum.org

Urban Institute www.urban.org

Welfare Information Network www.welfareinfo.org

W. E. Upjohn Institute for Employment Research www.upjohninst.org/uicenter.html

Australia

Australian Institute of Family Studies www.aifs.org.au

Commonwealth Department of Family and Community Services www.facs.gov.au

Melbourne Institute of Applied Economic and Social Research www.melbourneinstitute.com

Research School of Social Sciences, The Australian National University http://rsss.anu.edu.au

Social Policy Research Centre, University of New South Wales www.sprc.unsw.edu.au

United Kingdom

Care, Values and the Future of Welfare project, University of Leeds www.leeds.ac.uk/cava

Centre for Analysis of Social Exclusion, London School of Economics http://sticerd.lse.ac.uk/case

Centre for Research in Social Policy, Loughborough University www.lboro.ac.uk/departments/ss/centres/crsp

Economic and Social Research Council www.esrc.ac.uk

Department for Education and Skills (previously Department for Education and Employment) www.dfes.gov.uk

Department for Work and Pensions (previously Department of Social Security) www.dwp.gov.uk

Economic and Social Research Council project on Future of Work, University of Leeds www.leeds.ac.uk/esrcfutureofwork

Institute for Fiscal Studies www.ifs.org.uk

Institute for Labour Research, University of Essex www.essex.ac.uk/ilr

Institute for Social and Economic Research, University of Essex www.iser.essex.ac.uk

Joseph Rowntree Foundation www.jrf.org.uk

Policy Studies Institute www.psi.org.uk

Social Policy Research Unit www.york.ac.uk/inst/spru

Tavistock Institute www.tavinstitute.org

Canada

Applied Research Branch, Human Resources Development Canada www.hrdc-drhc.gc.ca/arb/arb-home.shtml

Canadian Policy Research Networks www.cprn.com

Canadian Public Policy http://economics.ca/cpp/en/index.php

Government of Canada publications www.Canada.gc.ca/publications/publi_e.html

Policy Research Initiative http://policyresearch.gc.ca/index_e.htm

Social Research and Demonstration Corporation www.srdc.org

1 Of course, beneficiaries can leave benefit through other means, eg by becoming ineligible through family transitions. This is particularly pertinent for lone parents. However, many of the barriers to employment identified here also have a negative impact on beneficiaries forming new and sustaining relationships. For example, research in the US suggests that lone mothers who take up employment may also improve their chances of finding a partner (McLaughlin and Lichter 1997). At the macro-level, labour markets and “marriage markets” intersect (Callister 1998). Concerns about benefit system effects on family transitions are related. For a review of welfare reform’s effects on non-marital childbearing and the formation of lone-parent families, see Lichter and Jayakody (2002).

2 See Hand (2001) for a description of the various welfare-reform policies and programmes that have been implemented and evaluated in the countries reviewed in this report and their effectiveness. We draw on the evaluation literature in this report only to identify potential barriers to employment. For example, if a policy to address the costs of childcare is found to be effective at increasing employment levels among the target population, this result serves as evidence that childcare costs represent a barrier to employment.

3 The term “family members” is used broadly to include, for example, unrelated individuals such as close friends who are in the care of the beneficiary.

4 See Hand (2002) for more comprehensive discussion and analysis of replacement rates and EMTRs in New Zealand.

ASSESSING BARRIERS TO CLINICAL TRIAL PARTICIPATION SAMPLE ASSESSMENT QUESTIONS
BARRIERS FACED BY DISABLED PEOPLE THE FOLLOWING ISSUES WERE
Barriers to Employment Among Longterm Beneficiaries a Review of

Tags: among long-term, levels among, among, employment, review, barriers, beneficiaries, longterm

BARRIERS TO EMPLOYMENT AMONG LONGTERM BENEFICIARIES A REVIEW OF