Radical
Statistics

The Journal

The Subjects

The Books

News

Links

About

Home

Low Pay Statistics: setting the scene

Keynote Speech from the Annual Conference of the Radical Statistics Group, London, February 2001.

Bharti Patel

Thank you for inviting the Low Pay Unit to address the 27th Annual Conference of the Radical Statistics Group.

Speaking at a conference of a group with the words statistics and radical in the title I am obliged to start with a disclaimer - I am neither a statistician nor a radical. However, my work at the Low Pay Unit, campaigning for a fairer deal for Britain's poorest workers has brought home to me the real value of statistics, not as an abstract concept, but in the real world where people's experiences of work and poverty are shaped by the decisions taken by politicians and civil servants - based on statistical evidence - so we are told!

As we approach the budget and the election period, statistics designed to give a measure of whether things have got better or worse off are used commonly by all political parties and their relevant support groups each claiming the record on success or failure of the different policy development.

As noted by a certain John Kennedy, not JFK, but a Texan civil servant:

If you torture statistics long enough they will confess to anything.

So who is torturing what statistics to make them confess to what claims...?

Figures on the Budget surplus, reported in the media, stretch from anything between £ 18bn to £ 40bn (including revenue from the third generation sale). Yet the Chancellor insists it will be £ 15bn.

When you talk about billions you lose 3/4 of the public, who are simply not able to picture such a sum. Such big numbers become meaningless. For example income support for a single pensioner totalling £ 72 per week or £ 3744 annually represents 0.000000208 of the surplus. But, the availability of extra £ 3bn or £ 25 bn can make a significant difference to what one requests from the Chancellor in the Budget.

For the Low Pay Unit the list would include better pay, better social security provision together with significant improvements in the public services.

Over-estimating or under-estimating figures results in poor policy development. The losers are those at the bottom of the income distribution who depend on public services and state intervention to lift them out of poverty.

The experience of the national minimum wage (NMW) illustrates this point. The NMW was set against the background of a freeze on public sector expenditure together with no increase in income tax rate and the threat of a projected one million job losses.

So when the Low Pay Commission (LPC) recommended the first level for the NMW for workers in the UK, it was set cautiously low. The Labour Government's caution exceeded that even of the LPC: the LPC recommended an initial adult rate of £ 3.60 for workers aged 21 and over and an initial development rate of £ 3.20 for 18 - 20 year olds, but the government set rates at £ 3.60 for adults aged 22 and over and £ 3 for 18 - 21 year old. This dramatically reduced its impact.

The NMW was originally intended by the LPC to cover around 2 million workers (10% of the workforce) but the number of people actually benefiting was estimated at 1.5m (more than 20% lower than originally planned). This figure was to eventually be 1.2m workers - just over half the number initially intended to benefit.

Table number of jobs paid at less than £ 3 per hour (aged18-21) or £ 3.60 per hour (aged 22 and over)

Source: ONS website:www.statistics.gov.uk

Previous estimate number below rate(millions)

New estimate number below rate (million) New Methodology

Spring 1998

1.9

1.5

Spring 1999

1.2

0.6

Spring2000

-

0.3

There are two reasons given for this: the economic slowdown, predicted by statisticians and economists, did not occur. Both wages and employment growth remained strong and the new methodology developed by the Office for National Statistics for estimating the numbers of low paid jobs is more accurate than the previous estimates, which were either under-estimated using the New Earnings Survey or over-estimated using the Labour Force Survey.

As the TUC pointed out in their submission to the LPC:

It is reasonable to conclude that if the information available now had been known at the time of LPC recommendation, LPC would have recommended a higher rate initially.

A million more workers lost out as a result of poor data. Do statistics really ever matter when it comes to decisions made at a political level?

Mr. Portillo, when employment secretary under John Major, insisted it would 'create a bureaucratic nightmare and cost at least a million jobs'. Two days as shadow chancellor he dropped his opposition and admitted it had caused less damage than the Tories feared.

Figures are often used by individuals to rebut particular policy developments. On paternity pay for fathers, which the government proposed to pay fathers two weeks at £ 100 a week, Britain's bosses screamed that it would cost the government £ 100m and businesses an absolute fortune in expensive replacement staff. This was a rather bogus statistic claimed The Independent:

There are 700,000 babies born in Britain in a year; not all fathers will take that time off; some already receive full wages If half were to take this leave, the cost would not more than £ 40m a year.

Irrespective of the numbers, the national minimum wage - together with other rights to paid holidays and rights to part time workers and improved maternity rights - have proved to be a success story for the new Labour government. The NMW enforcement team have successfully recovered £ 3 million in illegal underpayment.

Statistics and the Government record on poverty

There is now a range of data on the distribution of incomes of families and individuals. The most commonly used official datasets include: Household Below Average Income (HBAI - (derived from the Family Resource survey), Labour Force Survey, New Earnings Survey, Family Expenditure Survey, General Household Survey, and the British Household Panel Survey.

The government inherited a record number 14 million people living in poverty in the UK in 1997, of whom 4 million were children. Quite rightly both the Prime minister and the Chancellor pledged to eradicate child poverty. New Labour's commitment to the twin aims of rewarding work and eradicating child poverty with work being promoted as the best route out of poverty focuses on raising incomes of low paid families through tax credits. Tax credits can succeed in raising incomes in the short term, but they deal with tackling the consequences of poverty not the causes. The single biggest cause of poverty in the UK is low wages.

The treasury's documents claim that 1.2 million children have been lifted out of poverty since Labour came to office. Yet most of the official statistics are still showing the poor got poorer under Labour. These figures, as ministers are quick to point out, fail to take into account the major anti-poverty measures - National Minimum Wage, Working Families Tax Credit, increases in children's benefits and the minimum income guarantee for pensioners which were introduced after March 1999.

The problem is the data available are often out of date making it difficult to make a quick assessment of the full impact of policy. For example, the HBAI published in November 2000 are drawn from financial year ending March 1999. Thus they were already 18 months out of date and will be more than 2 years out of date by the election date of June 7th 2001.

Recent work done at the Microsimulation Unit in Cambridge by Holly Sutherland showed that 250,000 children in Britain's poorest households have become worse off. The Unit assessed the effect of all tax and benefit changes between 1997 and last year and the effect of the NMW on Britain's poorest families. The Observer headline claimed that Britain's "poorest children are getting poorer". As Mark Atikinson pointed out in The Guardian at the beginning of the year:

If unemployment statistics can be published within two months and the labour survey within three months why does it take 18 months to produce poverty figures. Even in arcane statistical circles, it seems the poor come at the bottom of the queue.

What we at the Low Pay Unit ask is for a more prompt response to collected facts.

Bharti Patel
Director
Low Pay Unit
bharti.patel@lowpayunit.org.uk

Home Page Top of page