Author
SMITH Noel; MIDDLETON Sue;
Title
A review of poverty dynamics research in the UK.
Journal citation/publication details
York: Joseph Rowntree Foundation, 2007. 114p.
Summary
This review of 115 reports makes clear that poverty dynamics (longitudinal) research provides ‘a dramatically more comprehensive understanding of poverty than point-in-time studies’. Although the evidence indicates that the poverty problem is significantly worse than the official statistics suggest, it also offers some good news – that most people’s experience of poverty is transient – and offers pointers to policy changes that could help reduce and protect against recurrent and persistent poverty.
Context
The Centre for Research in Social Policy at Loughborough University was commissioned by the Joseph Rowntree Foundation to examine the poverty dynamics research base ‘in order to take stock of this important development in the study of UK poverty.’
Methods
What sources were used?
The following databases and other electronic sources were searched between July and September 2004: ArticleFirst [via the database host OCLC); ASSIA (Applied Social Science Index and Abstracts); British Humanities Index; the British Library catalogue; Caredata [now Social Care Online]; PsycINFO; Regard [the former research database of the Economic and Social Research Council]; IBSS (International Bibliography of the Social Sciences); Index to Theses; Rlab ]the LSE Research Laboratory publications database]; Social Services Abstracts; UKOP (United Kingdom Official Publications); Web of Science; and ZETOC [British Library electronic tables of contents service to UK academia]. Thirty six websites, listed in Annex C, were hand searched between November 2004 and January 2005.
What search terms/strategies were used?
No search terms are given in the report, but the ‘proposed search strategy’ is available in the review protocol at http://www.crsp.ac.uk/projects/poverty_dynamics.html. It consists of a series of terms indicative of poverty, which were combined with a series of terms indicative of dynamics using the Boolean operator NEAR (in this case, meaning ‘within ten words of’). However, in the protocol the authors accept the need to develop specific strategies for each database because of differences in indexing practice, search interfaces etc. Thus the protocol may not reflect exactly what searching was done.
What criteria were used to decide on which studies to include?
Detailed inclusion and exclusion criteria are set out in Table 1. In general, the authors sought empirical studies of periods from 1946 onwards that relate to poverty in the UK, including comparative studies that include analysis of the UK. Studies could measure the poverty status of the same individuals, families or households over time (quantitatively or qualitatively), and were required to use indicators of income poverty, expenditure poverty, material deprivation or subjective poverty. Outcome measures used to measure poverty over time had to remain consistent.
Who decided on their relevance and quality?
The searches identified over 10,000 references, which were reduced to 1,320 after initial checking of titles and abstracts, and to 946 after removal of duplicates. These were all screened in full text against the inclusion criteria and a ‘pragmatic quality appraisal tool’. Quality criteria relating to the appropriateness of study designs, sample size, conduct of data collection and other methodological issues are listed in Table 2 but these were not used as the basis of a rigorous assessment for each criterion that might, for example, have involved contacting authors for further details. Studies were excluded if there was clear evidence that they contravened the criteria, and included if there was sufficient evidence to show that they were ‘good enough’ for review purposes.
How many studies were included and where were they from?
A total of 115 studies were reviewed: 62 reports (including government reports, conference papers and working papers); 25 journal papers; 20 book chapters; seven books; and one dissertation. Of these, the majority (85) were published between 2000 and 2004, and all but one were quantitative. They are listed in Annex A and all, by definition, were conducted in or relate to the UK. Over half used British Household Panel Survey data.
How were the study findings combined?
‘General thematic foci’ were developed for the review protocol and subsequently refined in the light of initial engagement with the literature. The thematic synthesis focuses on the findings within studies that were generated directly from longitudinal analyses, and does not cover the interpretations or policy recommendations derived from them.
Findings of the review
The review begins by examining different understandings of, or perspectives on, poverty: how it is measured, whether the perspective is static or dynamic, absolute or relative etc. This is followed by chapters on: the trigger events for entering poverty; the individual and household characteristics associated with entering poverty; the outcomes of these triggers and characteristics; and finally the triggers associated with the exit from poverty, and the factors preventing entry into it in the first place.
Dimensions of poverty
Poverty is a much more common experience than point-in-time research suggests. Dynamic (longtitudinal) research, in contrast, presents a more complex picture: for example, between 1991 and 1998 an average of 15% of the population were below the income poverty threshold at any one time, but 32% had fallen below that threshold at least once during the period. For most people, poverty is short-lived, but about a third experience recurrent poverty, leaving poverty and then re-entering it within a year because their ‘income mobility’ is short range. A minority, some 2% over the 1991-1998 period, experience persistent poverty, remaining under the threshold for the entire time.
Triggers of poverty
Poverty is more often triggered by changes in income (mainly loss of employment rather than a fall in income) than by changes in household composition. However, the relative importance of triggers varies by gender. For men, employment change is the dominant trigger, while for women both employment and family change are important, reflecting the fact that separation and divorce are more likely to trigger poverty for women than men. Retirement can also trigger poverty, but this is dependent on individual employment histories.
At-risk groups
Previous experience of poverty is a risk factor. Poverty in childhood increases the risk of poverty in adulthood, poverty in adulthood increases the risk of poverty in old age, and poverty in one generation increases the risk of poverty in the next. Education is the key factor in mediating poverty risk across the life course.
Age is a risk factor for children and older people, with the risks increasing for younger children and older pensioners. Gender is also a factor, with more women than men experiencing poverty. Among households, the risks are greater for single parent families with dependent children, larger families and those with younger children. The evidence on disability is contested, and it remains unclear whether poverty causes disability or vice versa. However, the link with unemployed status is clear. Unemployed people are about twice as likely as the general population to experience persistent poverty, but low paid and insecure employment also increases the risk: between 1994 and 1997 a third of working age adults living in poverty for two or more consecutive years were in work.
Escaping poverty
The most common events to trigger an exit from poverty are moving from unemployment to employment, followed by an increase in wages. Employment is the most effective means of keeping people out of poverty, but the evidence shows that poverty resistance is dependent not only on the number of employees in a household but also the degree to which they are in full time, sustained employment. For single parent families, some 70% of exits from poverty are associated with employment change, while 20% are linked to household change, including the movement to two-parent status. This movement is likely to enhance poverty resistance in the sense that household stability and continuity generally offer the greatest protection: the risks are lower for couple households who remain childless or do not increase their family size.
Authors' conclusions
Poverty dynamics research gives a more realistic and comprehensive understanding of UK poverty than do point-in-time studies. While they show that the scale of the problem is even greater than suggested in official figures, they may also enhance public support for the fight against poverty and suggest where effort may be most usefully concentrated. The review also identifies several gaps in the evidence base, including: 1) an understanding of how poverty dynamics vary according to the severity of poverty (virtually all research is based on a single income poverty threshold); 2) the types of poverty experienced by black and minority ethnic groups, and whether particular triggers and factors are particularly influential; 3) an understanding of the impact of educational attainment on poverty exit, as opposed to its role in protecting against poverty entry; and 4) an understanding of how the movement of individuals into and out of households affects poverty dynamics. The authors also make the case for a longitudinal qualitative research programme to explore trigger events and protective factors in more depth and to complement the findings of quantitative studies.
Implications for policy or practice
The evidence reviewed shows that most people who experience poverty leave it quite quickly, and it may be more useful for policy makers to focus on recurrent and persistent poverty: the current ‘one size fits all’ approach risks wasting resources on those who experience transient poverty, many of whom would probably leave it naturally without assistance. The extent of recurrent poverty raises critical questions about the degree to which movements above the official poverty line represent a genuine and sustainable exit, and whether a policy that focuses on poverty exit is the best use of resources: by failing to address ways of keeping people out of poverty once they have left, the government may risk devoting successive waves of resources to the same individuals as they drop below the threshold again and again.
The current focus on employment as the key route out of poverty is supported by the evidence. However, job loss is the single most significant factor in entry into poverty, and policies need to focus not simply on entry into employment but on job retention (to prevent recurrent poverty) and job progression (to help prevent in-work poverty). Sustained, progressive employment is crucial for escaping, and continuing to avoid, poverty year-to-year and across the life course. However, many employment histories include a mixture of periods in and out of work because of factors including unemployment, pregnancy and child care, illness, and the movement from two-parent to single-parent household status. Greater support for people at these trigger points could help to reduce recurrent poverty.