![Page 1: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/1.jpg)
The recruitment and retention of teachers in deprived schools
Rebecca AllenInstitute of Education, University of London
20th February 2013 – Presentation to CASE, LSE
![Page 2: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/2.jpg)
Overview of talk
1. Teacher turnover2. Estimates of turnover in X‐sectional data3. New hires in short panel4. Quitters in short panel5. Conclusion
![Page 3: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/3.jpg)
Teacher turnover
• Separation can be:– Layoffs (rare)– Retirements– Quits = go to work in another school, leave teaching (or private sector), leave labour force altogether
• Standard models of quits:– Pecuniary factors = wage offers, promotion or wage growth prospects
– Non‐pecuniary factors = characteristics of students, resources, ethos, management, collegiality
![Page 4: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/4.jpg)
Why might teacher turnover matter?
For children…• Turnover per se can be
harmful to student progress• High turnover schools need
to recruit novice teachers• High turnover might signal
general motivational problems at the school
For government budgets…• Most turnover is also
wastage out of the profession
• Teacher training costs government about £5.5k plus bursary
• Half of new trainees are no longer in state maintained sector after about 5 years
![Page 5: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/5.jpg)
Is teacher turnover a particular problem in disadvantaged schools?
• Existing UK evidence: Dolton and Newson(2003); Smithers and Robinson (2004)
• Strong US evidence, but very different circumstances
• Why?– Urban areas = denser teacher labour markets?– Young teachers like to teach in urban areas?– Tougher working environment?
![Page 6: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/6.jpg)
Teacher and school two‐sided matching models
• Schools have agreeable‐ness (A) and individuals have effectiveness (E)• If teachers and schools agree on the ranking of desirable schools and teachers, then market
equilibrium should see…– High E teachers matched with high A schools – may continue job search but at low intensity– Lower E teachers will tend to be matched with lower A schools, and will engage in higher intensity
job search
• If A and E poorly observed pre‐search…– High E teachers may accept a job at a low A school because it is better to search whilst employed– Low E teachers have a chance of a better A school, so that’s why you see greater quits out of low A
schools.
• Differential separations out of high and low A schools is likely to depend on…– the precision of pre‐hire measures of A and E– the degree of flexibility of wages– the degree of exogenous turnover
• Implies disadvantaged schools:– Hire teachers who continue their job search, almost regardless of experiences within school– Hire relatively ineffective teachers, on average– May induce further equilibrium effects – ineffective teachers = poor results = more families move
away from school
![Page 7: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/7.jpg)
School Workforce Census
• New annual census of all staff in schools from November 2010 onwards
• We know: teacher and school ID, age, sex, ethnicity, tenure in school and in contract, pay, FTE hours, role, highest qualification, degree subject and teaching subject
• We don’t know: length of service (pension records merger), whether they teach in private schools, training route, who they teach
• Missing observations not a huge problem; missingness on individual variables more serious
![Page 8: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/8.jpg)
Analysis of SWC today
Cross‐section of SWC 2010• Is there a correlation
between deprivation of wider neighbourhood and teacher turnover?
• What factors ‘account’ for this correlation?
Matched SWC 2010 and 2011• What do the new hires in
2011 look like?• Where do they come from?• How does this vary by
deprivation of school?• Profile of completed
tenures (quits)
![Page 9: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/9.jpg)
Inferring teacher turnover from cross‐sectional data
• We have a distribution of elapsed tenure (how long each teachers has been in their school so far)
• Renewal theory states relationship between elapsed and completed tenure:– Suppose completed tenure, τ, has pdf g(τ) with mean μ
– then pdf of elapsed tenure, ε, is f(ε) = [1‐G(ε)]/μ• If turnover process runs a long time, then expected completed tenure is equal to two times the elapsed tenure
![Page 10: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/10.jpg)
Inferring teacher turnover from cross‐sectional data
• We use tenure data to infer separation rates• Compare numbers of teachers with elapsed tenure of:– (0‐2 years) and (2‐4 years)– (0‐5 years) and (5‐10 years)
• In steady state these are separation rates• (2‐4 years – 0‐2 years)/(0‐2 years)
![Page 11: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/11.jpg)
Proportion of teachers with tenure of 0‐2 years
.3.3
5.4
.45
0 .2 .4 .6Neighbourhood IDACI
Age
<=30
Primary
.25
.3.3
5.4
.45
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Secondary
.12
.14
.16
.18
.2
0 .2 .4 .6Neighbourhood IDACI
30 <
Age
<=5
0
.12
.14
.16
.18
.2
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
.02
.04
.06
.08
.1
0 .2 .4 .6Neighbourhood IDACI
Age
> 5
0
.04.
06.0
8.1
.12
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Unit is a 2-percentile
![Page 12: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/12.jpg)
.1.1
5.2
.25
0 .2 .4 .6Neighbourhood IDACI
30 <
Age
<=5
0Primary
.1.1
5.2
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Secondary.3
5.4
.45
.5.5
5
0 .2 .4 .6Neighbourhood IDACI
Age
> 5
0
.3.4
.5.6
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Unit is a 2-percentile
Proportion of teachers with tenure of 10+ years
![Page 13: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/13.jpg)
Separation rate at 2 years
.3.3
5.4
.45
0 .2 .4 .6Neighbourhood IDACI
Age
<=3
0
Primary
.1.1
5.2
.25
.3
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Secondary
.3.3
5.4
.45
0 .2 .4 .6Neighbourhood IDACI
30 <
Age
<=5
0
.1.1
5.2
.25
.3
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
.3.3
5.4
.45
0 .2 .4 .6Neighbourhood IDACI
Age
> 5
0
.1.1
5.2
.25
.3
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Unit is a 2-percentile
![Page 14: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/14.jpg)
Separation rate at 5 years
.4.4
5.5
.55
0 .2 .4 .6Neighbourhood IDACI
Age
<=3
0
Primary
.3.4
.5.6
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Secondary
.4.4
5.5
.55
0 .2 .4 .6Neighbourhood IDACI
30 <
Age
<=5
0
.3.4
.5.6
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
.35
.4.4
5.5
.55
0 .2 .4 .6Neighbourhood IDACI
Age
> 5
0
.3.3
5.4
.45
.5.5
5
.1 .2 .3 .4 .5 .6Neighbourhood IDACI
Unit is a 2-percentile
![Page 15: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/15.jpg)
Conditional relationship between school disadvantage and turnover
Primary Secondary
Short tenure Long tenure Short tenure Long tenure
Base 16.38 ‐5.32 12.57 ‐26.70
+ pupil char. 12.87 ‐0.00 12.57 ‐21.36
+ market 7.60 3.72 12.02 ‐17.48
+ teacher 4.68 5.85 4.37 ‐8.25
Separation: (0‐2) | (2‐4) (0‐5) | (5‐10) (0‐2) | (2‐4) (0‐5) | (5‐10)
Base 4.94 12.96 16.88 20.51
+ pupil char. 4.32 10.27 23.13 18.97
+ market 4.01 4.40 20.00 17.18
+ teacher 3.09 ‐0.24 16.25 5.90
![Page 16: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/16.jpg)
Poor quality match between 2010 and 2011 School Workforce Census
New arrivals:77,785 teachers without same
school ID in 2010 (18%), of which
52,970 with tenure less than one year
(12%)
New arrivals:77,785 teachers without same
school ID in 2010 (18%), of which
52,970 with tenure less than one year
(12%)
434,846 primary and secondary teachers in
November 2011
434,846 primary and secondary teachers in
November 2011
88,648 teachers with ID in 2010 who are not in the same school in 2011, of which 65,342 are nowhere in 2011
88,648 teachers with ID in 2010 who are not in the same school in 2011, of which 65,342 are nowhere in 2011
![Page 17: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/17.jpg)
Characteristics of new hiresNew hires Stayers Within school diff.
Mean age 33 years 40 years 8 years younger
Age under 30 50% 18%
Age over 50 6% 24%
Female 74% 75% 2pp fewer women
White 86% 90% 3pp fewer white
Black 2% 2%
Asian 4% 3%
![Page 18: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/18.jpg)
Proportion new hires by school FSM
![Page 19: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/19.jpg)
New hires more likely to be brand new teachers in deprived schools
![Page 20: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/20.jpg)
Characteristics of new hires by deprivation of school
Most affluent quintile = 1
quintile = 2
quintile = 3
quintile = 4
Most deprived quintile = 5
Diff per 1 pp change in FSM
Within LA diff
Mean age 32.9 32.6 32.5 32.3 31.9 ‐0.0260 0.0094
Age under 30 48% 50% 50% 51% 53%
Age over 50 7% 7% 7% 7% 6%
Female 72% 74% 74% 74% 76% 0.0008 0.0017
White 89% 89% 87% 85% 81% ‐0.0023 ‐0.0007
Black 1% 1% 1% 2% 4%
Asian 2% 2% 3% 5% 6%
![Page 21: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/21.jpg)
Change in pay for movers
![Page 22: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/22.jpg)
Completed tenure for quitters
![Page 23: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/23.jpg)
Quit rate by tenure
![Page 24: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/24.jpg)
Logit School F.E. Logit
Tenure 1 to 3 y 0.55 0.49 0.55
Tenure 3 to 5 y 0.42 0.37 0.42
Tenure 5 to 10 y 0.33 0.27 0.33
Tenure 10 to 20 y 0.28 0.23 0.28
Tenure 20 to 40 y 0.34 0.30 0.34
Age 60s 3.22 3.81 3.25
Age 50s 1.33 1.41 1.33
Age 40s 0.79 0.77 0.80
Age 30s 0.83 0.81 0.83
Female*age60s 1.22 1.22 1.21
Female*age50s 1.06 1.09 1.07
Female*age40s 1.20 1.22 1.20
Female*age30s 1.13 1.13 1.13
Female 0.80 0.85 0.79
Ethnicity black 1.33 1.08 1.21
Ethnicity asian 1.12 1.02 1.05
% FSM 1.01
![Page 25: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/25.jpg)
Ofsted seems to think there is a ‘quality’ issue at deprived schools
2003 2004 2005 2006 2007 2008 2009
Excellent 8% 3% 18%
Outstanding/v. good 11% 12% 11% 6% 11% 12% 9%
Good 15% 14% 13% 11% 12% 11% 12%
Satisfactory 17% 17% 16% 14% 16% 14% 14%
Unsatisfactory 22% 17% 13% 19% 16% 18% 16%
Poor 26% 17% 22%
Very poor 16%
Note: Average free school meal eligibility for schools given judgment in year
![Page 26: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/26.jpg)
Less clear in recent GCSE data
![Page 27: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/27.jpg)
Correlation between % new hires and secondary school effectiveness
• For best 8 GCSE subjects (z‐score)– With pupil background controls +0.069– With pupil background controls and LA FX +0.089
• For average GCSE maths and English score– With pupil background controls n.s.– With pupil background controls and LA FX n.s.
• For between department (mat, eng, sci) pupil fixed effects – no correlation
![Page 28: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/28.jpg)
Summary
• Teacher turnover does appear to vary by neighbourhood deprivation of school
• Differences between affluent and deprived schools are not very large
• Largely accounted for by market and teacher characteristics– Either preferences of young teachers– Or low market attractiveness of disadvantaged schools
• Not clear how disruptive teacher turnover is to the performance of a school
![Page 29: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/29.jpg)
Outstanding research questions
• Can we develop a more sophisticated model of school and teacher matching?
• How do school experiences affect the propensity of teachers to move jobs or leave the profession?
• Do deprived schools face a different teacher labour market?– Quality of the applicant pool?– Quality of their teacher stock?
![Page 30: The recruitment and retention of teachers in deprived schools · 2013. 2. 25. · The recruitment and retention of teachers in deprived schools Rebecca Allen Institute of Education,](https://reader035.vdocument.in/reader035/viewer/2022070212/61085990533882258357c162/html5/thumbnails/30.jpg)
Answering these questions
• In School Workforce Census– Better data linkage, more variables– More years of data
• In IOE administrative records and a survey of new trainees– Full academic and career history– Personality, motivation, attitudes to teaching, stress
• Looking for natural randomisation – placement experiences on PGCE courses