Explaining exits from unemployment in the UK, 2006-09
Labour Force User Meeting - 2nd December 2010
Peter StamHousehold Labour Market and Developments Branch
Background
• Previous work
• Long (2009)
• Interest and development
• Objectives
• Data
LFS – Characteristics of individuals
Methodology – Heckman modelling
1. Modelling the probability of being unemployed
2. Modelling the length of unemployment
3. Modelling the probability of finding employment
LFS – Reference group
• Female• Aged 35 through 49• Living in “West Midlands Metropolitan County”• Not classified as an ethnic minority• Unmarried with no dependant children• Qualified to “Below GCSE”• Renting privately• Previous occupation “Elementary”
Modelling the probability of being unemployed
Marginal effect Statistical Significance
Age 18 through 24 4.5 ***
Age 25 through 34 - 1.1 ***
Age 50 through 59 - 2.0 ***
Age 60 plus - 4.5 ***
Male 3.2 ***
Ethnic minority 4.6 ***
Married - 5.5 ***
Dependent child and female 2.2 ***
GCSE 0.9 **
LFS - Results
LFS - Results
Modelling probability of having a spell of unemployment –Housing
-5%
-4%
-3%
-2%
-1%
0%
1%
2%
3%
4%
Rent-free Mortgage Own home Housing Association
LFS - Results
Modelling the length of unemployment Coefficient Statistical Significance
Constant 3.7 months **
Age 18 through 24 - 1.3 months *
Age 50 through 59 + 2.1 months **
Male - 1.6 months ***
Dependent child and male + 1.6 months ***
Job Seekers Allowance + 3.8 months ***
‘Administrative and Secretarial’ + 1.9 months *
‘Skilled Trades’ + 2.3 months **
‘Sales and Customer Service’ + 2.9 months **
‘Process, Plant and Machine’ + 2.6 months **
LFS – Results
Modelling the probability of finding employment
Marginal Effects
Statistical Significance
Male - 3.7 *
Married 9.1***
Dependent child and male - 1.7 *
Dependent child and female - 10.7 ***
Job Seekers Allowance 19.2*
GCSE 5.3 ***
Further Education 13.6***
Degree 12.1***
LFS - Results
-25%
-20%
-15%
-10%
-5%
0%
5%
10%
15%
Age 18 through 24 Age 25 through 34 Age 35 through 49 Age 50 through 59 Age 60 plus
Modelling the probability of finding employment –Age
LFS - Results
Spell length (months) Marginal effect
6 or less Base
7 - 12 - 17.7
13 - 18 - 27.2
19 - 24 - 11.3
25 - 30 - 21.0
37 - 42 - 12.1
49 - 54 - 24.3
60+ - 10.5
BHPS – Exit destinations
• Motivation• Economic states
• Data• Methodology
Inactive
?
Employed
Unemployed
BHPS - Results
Into employment
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47
Elapsed Duration (Months)
Haz
ard
BHPS - Results
Into inactivity
0.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47
Elapsed Duration (Months)
Haz
ard
Conclusions - Methods
• Modelling on the LFS provides consistent results with BHPS
• As duration of unemployment lengthens… so chance of re-employment decreases (LFS / BHPS)
• As duration of unemployment lengthens… so chance of inactivity increases(BHPS)
Conclusions - Messages
• Men more likely to become unemployed… but for a shorter time
• JSA increases the length of time unemployed… but increases the chance of finding employment
• Education has positive effect on finding employment (peaks at A-level)
• Housing association tenants more likely to be unemployed
• Mortgage holders increases chance of finding re-employment
• Older people have longer spells and are less likely to find re-employment
This work contains statistical data from ONS which is Crown copyright and reproduced with the permission of the controller of HMSO and Queen's Printer for Scotland. The use of the ONS statistical data in this work does not imply the endorsement of the ONS in relation to the interpretation or analysis of the statistical data. This work uses research datasets which may not exactly reproduce National Statistics aggregates.
Peter Stam
Household Labour Market and Developments
Telephone ext. 5982