The Application of Statistical Matching to the 2010 ESF Leavers Survey

Slides:



Advertisements
Similar presentations
Active labour market measures and entrepreneurship in Poland Rafał Trzciński Impact Evaluation Spring School Hungary,
Advertisements

Evaluation of Education Maintenance Allowance Pilots Sue Middleton - CRSP Carl Emmerson - IFS.
Estimating net impacts of the European Social Fund in England Paul Ainsworth Department for Work and Pensions July 2011
This research was supported by: U.S. Department of Education (U411B110098) and private-sector matched funds from 20 funders and foundations The Human Capital.
Welfare Reform and Lone Parents Employment in the UK Paul Gregg and Susan Harkness.
Evaluation of an ESF funded training program to firms: The Latvian case 1 Andrea Morescalchi Ministry of Finance, Riga (LV) March 2015 L. Elia, A.
Salford Futures 2013/14 Evaluation John Reehill Dave Timperley.
The Ogden Tables and Contingencies Other than Mortality Zoltan Butt Steven Haberman Richard Verrall Ogden Committee Meeting 21 July 2005.
ECE/ILO/Eurostat seminar on measurement of the quality of work (11-13 May 2005) The EU Labour Force Survey and indicators of quality in work.
Using propensity score matching to understand what works in reducing re-offending GSS Methodology Symposium Sarah French & Aidan Mews, Ministry of Justice.
What can a CIE tell us about the origins of negative treatment effects of a training programme Miroslav Štefánik miroslav.stefanik(at)savba.sk INCLUSIVE.
AGCAS Biennial Conference 2009, Brunel University, Uxbridge Thursday 10th September 2009 Gordon McKenzie.
Looking for statistical twins
Rhondda Cynon Taf County Borough Council
WIOA’s Goal Make Participants’ Skills Everyone’s Business
Overview of higher education statistics
How to rate success of a Business
ESF HEALTHY FUTURES PROJECT
Objective To have information to allow monitoring of SDG 8 from the point of view of persons with disabilities and to be able to provide guidance to policy.
Angelika H. Claussen, PhD,
L. Elia, A. Morescalchi, G. Santangelo
CSTL Sharing Meeting 2016 School Plus Programme
Evaluation of 15 projects – ‘Supporting School Leavers’
Background to the survey? Who were surveyed? My Voice findings
“CareerGuide for Schools”
Evaluation of the Employment Program Opportunity for All
LISA, Anticipating the Next Generation of Longitudinal Data
Employment and Social Affairs Platform
Propensity Score Matching Makes Program Evaluation Easy
LISA, Anticipating the Next Generation of Longitudinal Data
AES progress report and future plans
The European Statistical Training Programme (ESTP)
Challenges in Social Inclusion in Serbia
Adult Education Survey Implementation
Workshop on Measuring the Transition from School to Labour Market Item 3 – Conceptual framework in the EU for the transition of youth from education.
Orphaned Children Morrison and Ellwood (2000):
ESF EVALUATION PARTNERSHIP MEETING Bernhard Boockmann / Helmut Apel
Education and Training Statistics Working Group 24-25/9/2007
LAMAS Working Group June 2017
Impact Evaluation Methods
Econometric analysis of the benefits of early legal advice
Assessing Quality of Paradata to Better Understand the Data Collection Process for CAPI Social Surveys François Laflamme Milana Karaganis European Conference.
ESF Evaluation Partnership Meeting
Perpetrator Programs: What we know about completion and re-offending
2nd meeting of the task force on survey based disability statistics
Learning Seminar - Targeting employment policies
REAL (working with Singizi) Presentation to NSA 3rd/4th August
Resolution concerning statistics of Work, Employment & labour underutilization
Tuberculosis in Wales Annual Report 2018 Data to the end of 2017
LAMAS Working Group 29 June-1 July 2016
Meeting the emerging needs for statistics on migrants:
Measuring transition from School to Labour Market
Education and Training Statistics Working Group, May 2011
Metro ACEs Data 2018 Community Health Needs Assessment
Representative sampling Overview of the questions received by the ESF Data Support Centre Alphametrics Ltd. & Applica Sprl. Brussels, 13 March 2015.
Debriefing from the December 2017 LAMAS meeting Item 4
ESF Partnership meeting Marco Pompili – Ismeri Europa
Integrating Gender into Population and Housing Censuses
Evaluating the effects of ESF programmes
Counterfactual Impact Analysis applied in the ESF-Evaluation in Austria (period ) Contribution to the Expert-Hearing: Member States Experiences.
“Education and the labour market” in NewCronos
Andrew Jenkins and Rosalind Levačić
Data Management for FY2017 Reporting
Census topics selection
ESF Leavers Survey 2010: Use of Counterfactual Impact Evaluation
NEET – definitions and methodology
Mainstreaming essential For gender programmes For social programmes
Chapter 5: The analysis of nonresponse
Estimating net impacts of the European Social Fund in England
Tuberculosis in Wales Annual Report 2017 Data to the end of 2016
Presentation transcript:

The Application of Statistical Matching to the 2010 ESF Leavers Survey Rhys Davies Wales Institute for Social and Economic Research, Data and Methods, Cardiff University Gerry Makepeace Cardiff Business School, Cardiff University Findings are not in the public domain and are not for further dissemination

Overview of 2010 ESF Survey Aim – understand the characteristics and outcomes of those participating in ESF funded projects Methodology - telephone survey of people who were identified as having left an ESF project during 2010 Survey conducted Jun/July 2011 Interviews achieved with 7,509 participants (50% response rate) P2/P3 Convergence , P1/P2 Competitiveness Programme Similar survey conducted in 2010 among 2009 Leavers Previous and current activity, why did ESF, withdrawal from ESF, skills gained from ESF, educational attainment (pre/ESF/post), current employment, perceived additionality

Transitions in Main Activity: Results from the 2009 Survey Main activity before attending ESF course Current main activity at time of survey Paid employment Education and training Unemployed Economically inactive Total 8.3 1.5 1.1 0.6 11.4 5.6 3.9 3.0 1.3 13.8 21.8 7.4 21.0 6.1 56.4 2.8 2.6 10.7 17.5 38.7 14.3 28.0 19.9 100 Positive Transition No Transition Negative Transition

The Labour Force Survey as a Control Group ESF Survey provides no control group how do these transitions in to employment compare to those observed among wider population? what would ESF participants have done in the absence of ESF Use the UK Labour Force Survey to provide a control group against which the effect of participating in ESF can be evaluated Respondents to the LFS are asked what they were doing one year earlier so can look at transitions in economic activity over a period of 12 months Questions in ESF Survey and the coding of data are designed to align with definitions used in the LFS

Propensity Score Matching Use Propensity Score Matching to extract people from the Labour Force Survey who share similar characteristics to ESF participants Matching variables include age gender partnership status (couple/single/living at parental home), family status (dependent children under age of 18), health (work limiting health condition) educational attainment (NQF equivalents) ethnicity local employment (rate of employment among non-student population – allocated to decile groups)

Some Practical Issues The LFS data pre-dates the information from the ESF survey Introducing local labour market indicators requires access to detailed geographical identifiers Time elapsed between pre-ESF and current activity varies among ESF respondents – not set at 12 months a) the varying duration of ESF interventions and b) the different end dates OF these interventions Use career history data to identify activity 1 year following pre-ESF Only variables unaffected by participation within an ESF project should be used in matching – we use some variables that are measured at the time of the surveys Unemployed more homogenous than economically inactive

Career Profiles of Previously Unemployed – 2009 Survey

12 Month Transition Rates in to Employment Among the Previously Unemployed Local Area Employment Rates (non-student population of working age) LFS Transition Rates ESF Transition Rate (2009 Survey) 1st Decile (<68.8%) 29% 40% 2nd Decile (68.8-70.7%) 32% 3rd Decile (70.7-73.3%) 33% 4th Decile (73.3-74.4%) 34% 5th Decile (74.4-75.7%) 35% 6th Decile (75.7-76.6%) 42% 7th Decile (76.6-78.0%) 41% 8th Decile (78.0-79.5%) 44% 9th Decile (79.5-80.8%) 47% 10th Decile (>80.8%) 52% Total 38%

Propensity Score Matching Results Different matching techniques to consider sensitivity of results: Nearest Neighbour Select the individual who has ‘closest’ propensity score Radius Matching Average characteristics of controls within a specified radius With and without replacement allow controls to be matched to multiple ESF respondents Adjust size of callipers the maximum acceptable difference in propensity scores Include/exclude: ESF early withdrawers - Proxy respondents from the LFS Continuous career histories from the LFS

Conclusions, Limitations and Future Directions Additional insights can be provided by PSM if ESF surveys are designed to align with other available sources of data Limitations remain can only match individuals on the basis of measurable characteristics our control group may have participated in some other form of assistance some matching variables could have varied over the period during which employment transitions were being measured Future challenges refinement of statistical matching to use longitudinal data sources understanding of what works requires project level analysis