Characteristics of low-income households in the UK

Within the UK and across Europe the most common definition of low household income is one that is ⩽60% of the median household income in a given year. This measure is clearly relative, identifying a group whose level of income is socially unacceptable for the time and place in which they live⁽¹⁵⁾ and the threshold is used by civil servants, academics, politicians and researchers. Nevertheless, it does not indicate extents of income inequality, how rapidly such inequality might be changing or the length of time households spend in these circumstances, all of which can have profound effects on the capacity of households to manage their inevitable lack of essential goods and services. Thus, it represents a fairly crude indicator of poverty, which is why many surveys use a wider set of deprivation indicators (such as that used in the Low Income Diet and Nutrition Survey (LIDNS), see later) than household income alone. These issues are widely discussed in the social policy literature⁽Reference Gordon, Adelman and Ashworth^16–Reference Church, Babb, Martin and Haezewindt¹⁸⁾.

In 2005–6 there were approximately thirteen million individuals living in UK households whose income was ⩽60% of the median household income (22% of the population). The percentage in this category fell during the late 1990s and until 2004–5; about 15% of the population (nine million individuals) had been in this category for 2 of 3 years, i.e. living in what might be termed ‘persistent low-income poverty’. Many households also experience intermittent low income; repeated movements in and out of the category, since most do not suddenly become much richer but simply gain or lose a few pounds per week, which is termed ‘income churning’⁽¹⁹⁾. The most common reasons in the short term are rises in or losses of earnings by different household members or changes in receipt of state benefits; in the longer term, changes in household structure (arrival or departure of dependents, marital or partner separation) are more important, particularly for households headed by women⁽¹⁹⁾. Seemingly minor changes in available income can be critical in determining how much households can afford for food, which is an essential but flexible budget item, in that hunger can be assuaged quite cheaply and nutrients obtained from many different food commodities at very different costs⁽Reference Dowler, Turner and Dobson^20–Reference Dowler, Calvert, Köhler, Feichtinger, Barlösius and Dowler²²⁾.

UK households living on low income are typically those containing children or older adults, although, ‘over the last decade, the proportion of children and pensioners in poverty has fallen, while the proportion of working-age adults in poverty remained unchanged. As a result, ..... more than half the people now in poverty are working-age adults’⁽Reference Palmer, MacInnes and Kenway²³⁾. In fact, single working-age adults without dependent children (four million) are twice as likely to be poor as working-age couples without children. Adults with disabilities are also twice as likely to be in low-income households as non-disabled adults⁽²⁴⁾. Nevertheless, 3.8 million children live in low-income households, half of whom are in households in which one member is in paid work, and children are more likely to be living in low-income households than are adults, especially if they live with a lone parent. About one-sixth of those living on low incomes are pensioners (1.8 million), which is a lower rate than a decade ago, although many of those individuals entitled to state benefits to supplement their pension do not claim it. Low wages and job insecurity characterise those individuals in work who live in low-income households⁽²⁴⁾.

Evidence of problems: nutrition and health outcomes

Turning to food and nutritional outcomes, differentials in both household and individual dietary patterns, nutrient intakes and blood markers of nutritional status by various socio-economic indicators, in all age-groups and geographical regions, persist in the annual national surveys of household intakes (annual surveys of household food intake and expenditure published by the Ministry of Agriculture, Fisheries and Food until 2000, since then by the Department for Environment, Food and Rural Affairs⁽²⁵⁾) and in regular national surveys of individual intakes⁽Reference Gregory, Collins, Davies, Hughes and Clarke^26–Reference Hoare, Henderson, Bates, Prentice, Birch, Swan and Farron²⁹⁾. The importance of these observations for health and well-being is widely discussed⁽Reference Dowler, Turner and Dobson^20, Reference Acheson^30–Reference Press and Mwatsama³³⁾. The differences are particularly marked when intakes and food patterns are compared by household income, economic activity (employed v. unemployed and receiving state benefits) or household composition^(25, Reference Wrieden, Barton, Armstrong and McNeill³⁴⁾. Members of lone-parent households and those with two or more adults and children, which in the UK are more likely to be poor (however defined), are more likely to have lower micronutrient intakes than members of other household types⁽Reference Dowler^{21, 25)}.

The LIDNS recently published by the Food Standards Agency was a UK-wide survey specifically on the diet and nutrition of low-income households. Using a doorstep-administered questionnaire households who fell into (approximately) the bottom 15% of the population in terms of material deprivation were identified; the questions included receipt of benefits, household composition, car ownership and employment status⁽Reference Nelson, Ehrens, Bates, Church and Boshier³⁵⁾. The means of measuring food intake was based on intensive piloting of methods appropriate for households living in circumstances typical of those with low incomes⁽Reference Holmes, Dick and Nelson³⁶⁾. In general, the findings are similar to those of previous surveys, confirming that a considerable section of the population in question still fails to meet population dietary targets, has poor micronutrient intake, high BMI and blood pressure and low reported levels of physical activity. The food patterns are similar to those found in smaller-scale surveys, in that participants were more likely to consume high-fat processed foods or fast foods and snack foods, particularly children and younger adults, and less likely to eat the kinds of foods recommended for health, such as unsaturated-fat spreads and lower-fat milk, wholemeal products, vegetables and fruits and fewer sugary foods. Although average daily intakes of most vitamins and minerals from food sources were found to be above or close to the reference nutrient intakes, there was a proportion of the population whose intakes fell well below the reference nutrient intakes for Fe, Mg, K and Zn, and some were below the lower reference nutrient intake. For instance, about 48% of women <35 years, and about 52% of women aged 35–49 years and approximately 40% of girls aged 11–18 years were found to have Fe intakes below the lower reference nutrient intake⁽Reference Nelson, Ehrens, Bates, Church and Boshier³⁷⁾. LIDNS was the first to use a food-security questionnaire in a national survey. About 30% of households within this low-income population lived in food-insecure households, which meant that they had said that during the previous year their access to enough food that is both sufficiently varied and culturally appropriate to sustain an active and healthy life had been limited by lack of money or other resources. About 39% of participants reported having worried that their food would run out before money for more was obtained during the previous year and one-fifth reported that they reduced or skipped meals regularly because of lack of money⁽Reference Nelson, Ehrens, Bates, Church and Boshier³⁸⁾. In common with other survey findings, a higher proportion of this population were identified as current smokers (45% men and 40% women).

These national survey findings reflect the many small-scale, often qualitative, studies on the experience of living on a low income, some of which have specifically focused on food or on food insecurity⁽Reference Tingay, Tan, Tan, Tang, Teoh, Wong and Gulliford³⁹⁾, where individuals regularly report an inability to spend as much on food as they would wish or that having sufficient money is what limits their purchase of food they know to be healthier⁽Reference Hitchman, Christie, Harrison and Lang^40–Reference McKendrick, Cunningham-Burley and Backett-Milburn⁴²⁾.

Elevated BMI indicating overweight and obesity are also widely documented as socially patterned; women and, increasingly, men and children from lower socio-economic groups are more likely to be obese or overweight^(43–Reference Drewnowski and Specter⁴⁵⁾. The size of differentials depends on which social indicators are used (whether household income, education level, social class, receipt of means-tested benefits or area-deprivation scores) and childhood socio-economic conditions may contribute to the development of obesity over time⁽Reference Okasha, McCarron, McEwen, Durnin and Davey Smith⁴⁶⁾. The association between socio-economic status indicators and obesity also need to be interpreted with caution, as obesity may lead to lower educational attainment, lower employment and reduced income⁽Reference Lobstein and Millstone⁴⁷⁾. Nevertheless, household income and area deprivation seem to be better predictors of childhood obesity than a binary occupation-based indicator⁽Reference Stamatakis, Primatesta, Chinn, Rona and Flascheti^48–Reference Kinra, Baumer and Davey Smith⁵⁰⁾, the latter implying an ecological effect that has been observed for adults at country levels⁽Reference Pickett, Kelly, Brunner, Lobstein and Wilkinson⁵¹⁾, particularly for incidence of abdominal obesity⁽Reference Kahn, Tatham, Pamuk and Heath^52, Reference Ellaway, Anderson and Macintyre⁵³⁾. The recent UK Foresight Report on Trends and Drivers in Obesity has highlighted both the wide range of social, environmental, technological and other drivers of rising incidence of overweight and obesity, and has hinted that these factors can differentially affect those individuals in lower socio-economic groups⁽⁵⁴⁾. The need for a multi-sectoral policy that addresses the wider environment (including the workplace, transport etc.) as well as focusing on individual behaviours was particularly emphasised⁽⁵⁴⁾.

Interpreting the problem

These relationships highlight the challenges in interpretation; to what extent is low nutritional status a product of the material conditions of poverty and deprivation (that individuals do not have enough money to spend on appropriate food or cannot easily purchase it in the places where they live, work or go to school) rather than a result of particular circumstances of cultural and social capital, which may include aspects of social status and may more probably be indicated by educational levels or occupationally-based social class?⁽Reference Dowler, Caraher, Lincoln, Dowler and Spencer^4, Reference Crotty, Germov, Germov and Williams^55, Reference Dahl, Fritzell, Lahelma, Martikainene, Kunst, Mackenbach, Siegrist and Marmot⁵⁶⁾. Social capital, the ‘resources (material, social and in-kind) available to individuals through their social behaviours and membership in community networks’⁽Reference Kawachi⁵⁷⁾, has recently been examined in relation to food and has been shown to have a positive link with food security and a higher social capital leads to a lower risk of hunger at the household level in the USA⁽Reference Martin, Rogers, Cook and Joseph⁵⁸⁾. European attention to social exclusion, which recognises multidimensional disadvantages for individuals and areas and its role in perpetuating deprivation, has informed recent UK approaches to reducing inequalities (although not specifically related to food), not least in the work of the Social Exclusion Unit and latterly the Cabinet Office Social Exclusion Task Force⁽⁵⁹⁾. The reliance in much contemporary analysis of food patterns and nutritional indices on indicators of socio-economic status such as occupation of household head or education level as proxies for the material conditions of the household (signalled by ‘income’) are likely to mask both complexities of social differentiation in a society such as the UK (of gender, ethnicity, religion, age, area and community, which are distinct but linked) and also household circumstances alluded to earlier (for how long individuals have been living in deprivation, why, and what characterises their living conditions)⁽Reference Graham and Kelly⁶⁰⁾. There are further complexities to teasing out the relationships between food and nutritional outcomes and factors such as money, skills, cultural or social capital, in that the effects might vary depending on whether they pertain to the individual or to the household; intra-household management of, and access to, resources of various kinds may not be equal, and this factor may be more important for outcomes in some circumstances⁽Reference Goode, Callender and Lister^61, Reference Lister⁶²⁾. Moreover, the relationship between food and nutritional outcomes and indicators of area deprivation may be different in different places and/or times.

These issues of definition and conceptualisation are critical to developing improved understanding of the nature of the problems involved and the framing of appropriate responses; they affect targeting and intervention by the state or other agencies at national and local levels. Beyond these decisions there is also the question of whether what matters are those who live on low(est) incomes (i.e. below a minimum subsistence level or that needed to meet minimum requirements for health⁽Reference Morris, Donkin, Wonderling, Wilkinson and Dowler⁶³⁾: what is described as ‘absolute poverty’⁽Reference Dowler and Dobson¹⁴⁾) or those living on household incomes in the lowest deciles or quintiles of national income distribution or other deprivation indices (as in LIDNS described earlier or in the UK poverty statistics of ‘households below average income’⁽Reference Palmer, MacInnes and Kenway²³⁾). Reducing inequalities is not the same as improving the nutrition and health of those individuals who are worst off in society. In the field of health inequalities there has been considerable work on the theoretical and practical implications of these different approaches and much debate on the salience of social capital and exclusion⁽Reference Lynch, Smith, Kaplan and House^64–Reference Baum and Ziersch⁶⁷⁾. As yet, this work has not been matched by comparable research within the more specialised fields of nutrition and diet. Nevertheless, the major survey of low income and diet in the UK, the LIDNS report, has been promoted as showing that low-income households are not a distinguishable group with considerably worse nutrition than the rest of the population. It was argued, in the press release and subsequently, that since many UK households do not meet current dietary guidelines, low-income households are no different and do not need special attention⁽⁶⁸⁾. Whether or not this argument strengthens the need for interventions that simply focus on changing the dietary pattern of the majority of the UK population, on the expectation that any benefit would include or even ‘trickle down’ to low-income households, rather than more targeted intervention in social or nutrition policy for those on low income (however defined), is an ongoing critical discussion at present (T Lobstein, unpublished results).

In other words, to what extent do problems reside in the particular circumstances and conditions of specific groups, so requiring intervention aimed at them, or is it that problems are more generalised such that the drivers of the outcomes observed affect so many within the population that more general policy approaches are required? The latter does not preclude attention to the needs of specific groups (to ensure their capacity to capture the benefits), but the thinking and approach is different from one that targets by income, area deprivation or some other individual, household or locality indicator. Of course, it may be politically difficult to identify specific groups as experiencing nutritional problems and requiring targeted solutions (whether because of social or racial prejudice, or potential stigmatising).

Evidence of causal understanding: inputs and choice

The earlier caveats notwithstanding, there have been a number of studies trying to elucidate the determinants of the nutritional situation of those individuals who are poor and/or living on low incomes, addressing both material conditions and personal choices and practices. On the structural side, the UK, in common with the USA and Australia, has seen research on access to healthy food choices, with particular interest in the distance needing to be travelled (and means of doing so) to reach shops selling a good range of foodstuffs⁽Reference Donkin, Dowler, Stevenson and Turner^69–73). In Scotland a national map of access to healthy food was commissioned⁽⁷⁴⁾ and the salience of ‘food deserts’ (crudely, areas in which there are few shops selling sufficient healthy food) has been much debated⁽Reference Cummins and Macintyre^75–Reference White, Bunting, Williams, Raybould, Adamson and Mathers⁷⁷⁾. There has been rather less research on other aspects of access, such as how money is managed. Basic expenditures such as rent, fuel and water can absorb a high proportion of outgoings and have the force of mandatory collection; such costs have risen faster than the retail price index in the UK in recent years, and they vary around the country, unlike income from benefits, pensions or the minimum wage⁽Reference Kempson and Bennett^78–81). The expenditure needs of children, and the cost of food relative to other essentials, can also be very critical in determining purchasing patterns⁽Reference Dowler, Dowler and Jones Finer⁸²⁾, especially when low household income can fluctuate because of factors outside members' control. The cost of food can also vary between shops (and around the country), even for the same commodities. Thus, individuals living in different household circumstances may face very different constraints on how much money they can allocate to food.

The more theoretical approach of budget standards research, which cost diets appropriate for health for different household types, has demonstrated that, on average, UK families living on low incomes with dependent children⁽Reference Parker⁸³⁾, young single men paid at or below the minimum wage⁽Reference Morris, Donkin, Wonderling, Wilkinson and Dowler⁸⁴⁾ and older individuals living on state old-age pensions⁽Reference Morris, Dangour, Deeming, Fletcher and Wilkinson⁸⁵⁾ are likely to have insufficient money to meet basic needs for healthy living, including food. In that they do not, on the whole, go hungry is testament to immense skill in budgeting and managing limited resources, acceptance of monotonous diets and the probability of having to rely on very cheap food that is unlikely to contribute to a healthy dietary intake. It also offers explanation for the LIDNS findings on food security summarised earlier.

Nevertheless, money is not the only determinant of food purchase; skills, taste and culture clearly do shape what individuals buy and how and what they eat, although of course these determinants are not divorced from the material aspects of life. It has been argued that these factors are more important in shaping an individual's purchases than structural aspects of access, such as distance to shops and what is available in them⁽Reference Cummins and Macintyre^75, Reference White, Bunting, Williams, Raybould, Adamson and Mathers^77, Reference Dibsdall, Lambert, Bobbin and Frewer⁸⁶⁾. When asked through surveys, individuals are shown to rate personal likes, or beliefs over what is appropriate to eat, as more important in their decision making than commodity availability or physical access to large supermarkets⁽Reference White, Bunting, Williams, Raybould, Adamson and Mathers^77, Reference Dibsdall, Lambert, Bobbin and Frewer⁸⁶⁾. Such responses may reflect methodological approaches that ask individuals to rank aspects governing choice in a questionnaire; different answers are sometimes obtained in qualitative inquiry⁽Reference Beresford, Green, Lister and Woodard^87–Reference Kyle and Blair⁸⁹⁾. This disparity is partly because the characteristics of the physical and social environment in which individuals live or work may influence and frame the largely unconscious everyday practices of food purchase and consumption, so that ‘availability and access’ are not ‘perceived’ as issues, whether relating to the type and location of shops and markets or the provision in schools or workplace canteens. It is nevertheless useful to try to disentangle, for specific communities in any given place, how the complex set of social conditions and practices condition food choice among low-income consumers, and thus dietary and nutritional outcomes, and how different factors may jostle for supremacy day-by-day or from week-to-week. Such understanding can then shape intervention at local level.

Finally, it should be noted that in the UK there has been rather less research on the food and nutritional experience of those individuals whose circumstances fall outside official surveys and responses, particularly those who are roofless, homeless and/or asylum seekers. The few studies that have been done demonstrate that conditions and outcomes in terms of food patterns and nutrient intakes are worse than for those living in accommodation to which they have entitlement, those who have citizenship status and those whose lives are less fragmented and difficult⁽Reference Evans^90–Reference Sellen, Tedstone and Frize⁹⁵⁾.

Policy responses

Responses by the state to poor nutritional outcomes in low-income households have varied over the last century⁽Reference Smith⁹⁶⁾. Most recently, in the 1990s in England, the Department of Health has led in setting up a Project Team on low income and diet under the Nutrition Task Force (1993–6). The subsequent report addressed structural and material issues, such as changes in food retailing and the responsibilities of local and national government^(97, Reference Nelson⁹⁸⁾, calling for a national network of local initiatives on food and low income and the creation of local public–private sector food partnerships, especially in areas of multiple disadvantage, to regenerate local food economies. Fig. 1, which is modified from a version produced for the Project Team, shows the range of determinants of nutritional intakes that were considered. These ideas informed both the Acheson Inquiry into Inequalities in Health⁽Reference Acheson³⁰⁾ and the early public health White Paper under the new Labour Government (elected in May 1997 with a mandate to reduce inequalities)⁽⁹⁹⁾. The Acheson Report includes within its many recommendations increasing the availability and accessibility of an adequate affordable diet, mentioning the need for policies to ensure adequate retail provision of food to those individuals who are disadvantaged. There is also a call for policies to reduce income inequalities and improve the living standards of households in receipt of social security benefits. The recommendations are linked to policies to improve the health and nutrition of women of child-bearing age and their children, with priority given to the elimination of food poverty and the prevention and reduction of obesity⁽Reference Spencer, Dowler, Dowler and Spencer¹⁰⁰⁾.

Fig. 1. Main determinants of food and nutritional intake by households and individuals in developed countries. (Adapted from Department of Health⁽⁹⁷⁾, Dowler et al. ⁽Reference Dowler, Turner and Dobson²⁰⁾ and Robertson et al. ⁽Reference Robertson¹⁴⁷⁾.)

Since the publication of the Acheson Report a number and range of policy initiatives have been set in train, if not fully implemented. Although in practice the UK government and its devolved institutions have not explicitly addressed the issues in inequalities when applied to food and nutrition discussed earlier and, in fact, embrace most approaches, ‘food and nutrition’ has been located as a policy issue in different places at different times, particularly when seen as addressing the needs and priorities of low-income households. Indeed, the complexities of Fig. 1 make clear how much the determinants of dietary intake depend on a range of factors and policy arenas. The Department of Health traditionally led on such initiatives and has continued to do so over the last decade, partly because of a focus on health inequalities. Nevertheless, the former Ministry of Agriculture, Fisheries and Food commissioned research on the needs of low-income households and a scoping study on challenges of surveying their circumstances and conditions of food and nutrition⁽Reference Dowler, Draper, Thomas, Dobson and Nelson¹⁰¹⁾. This role has now been taken over by the Food Standards Agency, the independent Government department set up in 2000 to protect the public's health and consumer interests in relation to food; the Food Standards Agency commissioned the LIDNS discussed earlier⁽Reference Nelson, Ehrens, Bates, Church and Boshier³⁵⁾. However, at present neither the Food Standards Agency nor any other body coordinates policy focus on low-income households' nutritional needs. Indeed, it is quite difficult to disentangle sectoral responsibilities for aspects of food and nutrition since they have been shifting over recent years and differ in the devolved territories. The following comments are presented broadly under topics to help clarity, but the boundaries between sectors or even substantive areas are not consistent and there are overlapping responsibilities and practices.