The New Policy Institute has produced its fifth annual report of indicators of poverty and social exclusion. The data is the most comprehensive and up-to-date available.

• In 2000/01, there were 13 million people living in relative low-income households. This is a fall of 1 million - or 7 per cent - since 1996/97. The falls thus far have only been sufficient to bring the numbers back to the 1995/96 levels and are still almost double those of twenty years ago.
• Three-fifths of people wanting paid work are economically inactive - chiefly lone parents and those who are sick and disabled - compared with two-fifths who are officially (ILO) unemployed. This is a reversal of the position from a decade ago.
• Low pay remains prevalent, with an estimated 2 million adults paid at or just above the national minimum wage. Those in low-paid jobs are very unlikely to get any work-related training.
• There were substantial reductions in the number of pupils failing to achieve basic education qualifications throughout the 1990s. For example, a quarter of 16-year-olds failed to achieve any GCSEs above a D in 2001 compared with a third in the mid-1990s.
• Levels of obesity and treatment for drug misuse have been rising. Accidental deaths among children, youth suicide and underage pregnancies have been falling. In all cases, lower socio-economic groups are affected disproportionately.
• Many fewer households, including poorer households, now lack central heating than five years ago.
• The number of burglaries continues to decline, down to almost half of its peak in the early 1990s.
• The number of households in temporary accommodation has almost doubled over the last five years.
• Despite government initiatives, a fifth of the poorest households still remain without any type of bank or building society account.
• Half of all low-income households still lack household contents insurance. Households without insurance are three times more likely to be burgled than those with insurance.
• Two-thirds of heads of households in social housing do not have paid work, the same as a decade ago.

The pattern for the number of children in low-income households is similar, with 3.9 million children living in households below the low-income threshold in 2000/01. This is a fall of 200,000 compared with 1999/00 and a fall of 500,000 since the peak in 1996/97, but is the same number as at the start of the 1990s. Children remain much more likely to be in low-income households than adults.

The real incomes of the poorest increased by about 1.5 per cent per annum (£2 per week) in real terms during the 1990s. These increases are reflected by the fall in the numbers below the fixed threshold of 60 per cent of 1994/95 median income, from 13 million in 1994/95 to 8.5 million in 2000/01. The real incomes of the poorest and at the median increased at roughly the same percentage rate during the 1990s, and this is why the numbers on relative low incomes remained largely unchanged.

... and the government's 2004 target represents a major challenge ...
The government has set a specific target to reduce the number of children in low-income households by at least a quarter by 2004/05 compared with 1998/99.

Simple arithmetic suggests that the government's target is potentially achievable: the 2000/01 figures were around a third of the way through the period and around a third of the target had been achieved. But further progress cannot be taken for granted. The initial impact of the national minimum wage and Working Families Tax Credit has already been included in the 2000/01 figures. Furthermore, the main reason that the numbers have fallen in recent years is because the numbers of unemployed (a high risk group) fell and the numbers of employed (a low risk group) rose, rather than because the risk of low income of either group changed substantially. With unemployment at a historic low, there appears to be limited scope for further falls of this type.

One of the issues here is the continuing large number of people who are economically inactive but want paid work, many of whom are 'sick and disabled'. At 2.5 million people, these now substantially outnumber the number of people classed as unemployed by the ILO because they are actively seeking work (now down to 1.5 million people) (see Figure 2). Many sick and disabled people find it difficult to obtain paid work and this clearly poses a serious challenge to the strategy of reducing poverty by encouraging people into work. The scale of this challenge is reinforced by the lack of change in the number of long-term workless households, which has remained at about 2 million since 1996.

... with disadvantage at work continuing to be a major problem.
Since the introduction of the national minimum wage in 1998, the official estimate of the number of adults paid below this hourly rate of pay has dropped dramatically from 1.5 million in 1998 to 200,000 in 2001. But the trend is much less favourable when looking at slightly higher rates of hourly pay: in 1998, 2.5 million adults aged 22 or over were paid less than £4 per hour; by 2001, an estimated 2 million were still being paid less than this hourly rate, after adjusting for inflation. The vast majority of these are women and represent a particularly high proportion of the workforce in the distribution, hotels and restaurant trades.

This continuing prevalence of low pay is illustrated by the observation that half of all children and working-age adults living in low-income households have someone in their household in paid work. This connection reinforces concerns about the lack of adequate official data about the extent of low pay, which is now only available at government discretion, with a substantial time lag and subject to a complex statistical procedure to derive the 'estimates'.

Disadvantage at work is not limited to low pay. Those without any educational qualifications are three times less likely to get work-related training as those with qualifications. And levels of job insecurity are illustrated by the fact that around two-fifths of people who make a new claim for Jobseeker's Allowance last claimed less than six months ago, the same as in 1995.

Favourable trends in education ...
The general pattern here is of substantial improvements throughout much of the 1990s. For example, a quarter of 16-year-olds failed to achieve any GCSEs above a D in 2001 compared with a third in 1994 (although unchanged since 1999). A quarter of 11-year-olds failed to achieve Level 4 at Key Stage 2 in English in 2001 compared with more than two-fifths in 1996 (with a similar pattern in maths). And a quarter of 19-year-olds were without a basic qualification in 2001 compared with a third in 1992 (see Figure 3).

For 11-year-olds at least, it appears that schools in more deprived areas are fully sharing in the improved educational outcomes: two-fifths of children in schools with high numbers of children on free school meals failed to achieve Level 4 at Key Stage 2 in English and maths in 2001 compared with three-fifths in 1996. These schools' results in 2001 were actually better than the all-schools average in 1996.

Even though the trends are favourable, low attainment at school remains a serious problem, with 150,000 19-year-olds lacking a basic qualification (at least NVQ2 or equivalent) and 60,000 (4 per cent of the age group) having no qualifications are all (at least GCSEs at grade G or above).

... in housing ...
Less than 15 per cent of poorer families now lack central heating, compared with 25 per cent in 1994/95 (see Figure 4). The scale of progress is illustrated by the fact that a greater proportion of poorer families now have central heating than did the population as a whole in 1994/95. Notwithstanding an apparent rise in the latest statistics (2000), the proportion of households which are overcrowded is still down by a third compared with a decade ago.

... and in crime.
The number of burglaries continues to decline, down to almost half of its peak in 1993 and lower than at any time in the previous 15 years. Lone parents and households headed by young people (aged 16 to 24) are still three times more likely to be burgled than the average, and low-income households one and a half times more likely.

But a doubling in those in temporary accommodation ...
The number of households in temporary accommodation (a much wider group than rough sleepers) continues to rise and has now almost doubled over the last five years, from 45,000 in 1997 to 85,000 in 2002.

... and no progress in financial exclusion ...
In 2000/01, one in five of the poorest households did not have any type of bank or building society account, the same proportion as in 1994/95. This compares with one in twenty households on average incomes (see Figure 5). The lack of change is despite government initiatives to encourage the development of bank accounts for people who are not comfortable with the products currently on the market.

Half of all low-income households still lack household contents insurance, the same as five years ago. This compares with a fifth of households on average income. Yet households with no household insurance are around three times as likely to be burgled as those with insurance and the impact is obviously particularly serious for those on low incomes who, by definition, are less able to replace stolen goods themselves.

... or in access to work by those in social housing.
Two-thirds of heads of households in social housing do not have paid work, the same as a decade ago and twice the rate in other tenures. This difference between tenures is reflected in the relatively low incomes of social housing tenants, with two-thirds living on a weekly income of less than £200 compared with a quarter of those living in other tenures.

Reflecting these differences, people in low-income households are twice as likely to report that their quality of life is significantly affected by fear of crime than the average and one and a half times as likely to feel very dissatisfied with the area in which they live.

An absence of up-to-date statistics on health.
Although quite a lot of data is available on various aspects of health, most of it is somewhat out-of-date, with 2000 typically being the latest year available. In part, this is because there appears to be some delay in the government processing of the 2001 data in some areas - notably premature death and accidental deaths.

The 2001 data on premature deaths will be important to see if the significant fall in geographic concentrations of premature death which occurred in 2000 represents the start of a downward trend or a blip in the statistics (see Figure 6). The number of accidental deaths of children halved during the 1990s and, furthermore, it appears that the rate of reduction was greatest in children from manual backgrounds.

The number of births to girls conceiving before their 16th birthday fell by a fifth between 1996 and 2000, but the vast majority of these births were still concentrated in those manual backgrounds. The number of suicides amongst 15- to 24-year-olds in England and Wales has also been declining since 1997.

In contrast, the proportion of babies who are of low birthweight did not change between 1995 and 2000, the level of young people starting treatment for problem drug use each year was double the levels of the early 1990s, and levels of obesity are at an historic high.