APPENDIX 11
Memorandum from Anne Green, Principal
Research Fellow, Institute for Employment Research, University
of Warwick (JG 17)
1. INTRODUCTION
This submission is based mainly on evidence
from the following publications reporting selected results of
research funded by the Joseph Rowntree Foundation and the Economic
and Social Research Council:
(ii) Geographical variations in unemployment
and non-employment.
(iii) Green A E and Hasluck C (1998) "(Non)
Participation in the labour market: alternative indicators and
estimates of labour reserve in United Kingdom regions", Environment
and Planning A 30, 543-558.
(iv) Green A E (1998) "Problems of measuring
participation in the labour market" in Dorling D and Simpson
S (eds) Statistics in Society. London: Arnold. pp 313-323.
2. ISSUE ADDRESSED
The key findings from this body of research
relate primarily to the question addressing:
— how successful the official measures,
such as the claimant count area statistics and those provided
by the Labour Force Survey, are at presenting the spatial disparity
in UK unemployment.
The main points arising are summarised below.
3. MEASURING
UNEMPLOYMENT: CONVENTIONAL
APPROACHES
The task of defining and measuring unemployment
in a clear and unambiguous fashion is a difficult one. In labour
market terms the adult population is conventionally divided into
three main categories (on the basis of ILO definitions):
— the employed: those who have a paid
job (of at least one hour's duration) in an employee or self-employed
capacity, or are on government-supported training and employment
programmes, or are unpaid family workers;
— the unemployed: those who do not
have a job but are actively seeking work and are available to
take up a job;
— the inactive: all remaining members
of the population.
The employed and unemployed together comprise
the economically active. The unemployment rate is conventionally
defined as the unemployed expressed as a percentage of the economically
active. The unemployed and inactive together comprise the non-employed.
4. CONVENTIONAL
APPROACHES IN
THE CONTEXT
OF LABOUR
MARKET DEVELOPMENTS
Over recent years there have been a number of
important changes in the UK labour market, which in turn have
prompted some users of labour market statistics to question the
continuing applicability of conventional approaches to measuring
participation, and the usefulness of the standard statistics outlined
above. Key features of labour market restructuring include:
— the demise of jobs in manufacturing
and the growth of employment in services;
— a reduced demand for traditional
skilled manual labour—predominantly men;
— a greater premium on higher level
skills and qualifications—and concomitantly, a reduction
in employment opportunities for those with no or few formal qualifications;
— a growth in flexible working and
labour market insecurity;
— a greater number of women in employment;
and
— the entrenchment of high levels of
unemployment and non-participation—particularly amongst
some sub-groups of the population and in some areas.
The result of these trends is a greater variety
of patterns of work and non-work—with more individuals in
irregular jobs and with discontinuous employment patterns, culminating
in a "blurring of boundaries" between employment, unemployment
and inactivity. The labour market of the 1990s has been variously
characterised as "fuzzy", "complex" and "fluid":
there are all kinds of "grey areas" on the fringes of
employment, unemployment and inactivity as conventionally defined.
5. STATISTICAL
ISSUES
Particular concerns have been levelled about
the validity of unemployment statistics, and the applicability
of conventional approaches to measuring unemployment in the face
of the changes in the labour market outlined above. Reviews of
statistics (such as those by the Royal Statistical Society and
the House of Commons Employment Committee) encompassed debates
on the relative merits of different data sources, issues of definition
and issues of interpretation.
6. THE CASE
FOR BROADER
MEASURES OF
UNEMPLOYMENT AND
NON-EMPLOYMENT
It may be contended that the complexity of the
labour market in 1990s is such that a wider range of statistics
than those conventionally used are needed to capture the experience
of labour market participation and non-participation. One group
of researchers has focused their efforts on generating a count
of "real unemployment" to set alongside the "headline"
claimant count measure at a local level. A more flexible approach
may be to manipulate available data sources to generate a suite
of cumulative measures of unemployment and non-employment—in
which sub-groups of the inactive are added to the unemployed to
generate broader measures of unemployment (and non-employment).
7. INTER-REGIONAL
DIFFERENTIALS
At the regional level the geography of unemployment
in the UK from the mid 1980s to the mid 1990s was characterised
by lower than average unemployment rates throughout the period
in the Rest of the South East, East Anglia, the South West and
the East Midlands. At the same time Northern Ireland, the Northern
region, the North West, Wales, Scotland, Yorkshire and Humberside
and the West Midlands have displayed generally higher than national
average unemployment rates—although a convergence to the
national average was evident in the early 1990s recession. Greater
London emerges as distinctive in moving from a position of lower
than average to higher than average unemployment rates over the
period. There are important regional differentials in the extent
of broader "unemployment" and "non-employment"
not captured by the conventional measure of ILO unemployment.
In general, compared with the picture for the UK as a whole, the
ILO unemployment measure captures a relatively larger proportion
of those included in the broader measures of "non-employment"
in the southern regions of England (including London), than in
Wales and the northern regions of the UK. This pattern indicates
that in regions characterised by long-standing labour market disadvantage
non-participation tends to be more entrenched, and the "pressures
causing a diversion from recorded to hidden unemployment are likely
to be greatest". This is in line with the general rule that
"the greater the degree of labour market disadvantage the
less appropriate is unemployment as a measure of labour market
slack."
8. LOCAL VARIATIONS
IN UNEMPLOYMENT
AND NON-EMPLOYMENT
Analyses conducted at the local level confirm
the findings evident at regional level that inactivity rates are
disproportionately higher in "high unemployment" areas
than in "low unemployment" areas and that there are
important spatial variations in the pattern of unemployment:non-employment
differentials. At the local level increases in activity rates
for men since 1981 have been most pronounced in mining and industrial
areas, while at the micro-area level neighbourhoods in inner city
areas and concentrations of public sector housing have witnessed
the largest increases in non-employment. Moreover, analyses of
data on off-flows from the claimant count reveal that the proportion
of those entering work on leaving unemployment in such areas was
lower than average.
Anne E Green
IER, University of Warwick
7 October 1999
| 
