Quality Of Education (Chaudhury , Muralidharan) Flashcards
What is the key thing to understand about quality of eduction
Just because people are enrolling, it doesn’t mean they are learning.
E.g in India 76% of grade 4 children cannot read do simple division.
So quality of education matters, not just enrolment
Is a lack of inputs e.g textbooks, desks etc a problem
No. Kremer and Holla found more inputs by themselves do not make a difference
Is the teaching the problem? Find 2 evidences for yes in Chaudhury et al
Yes -
High absenteeism, Found teachers miss 1/5 days in 2002/03 in Bangladesh etc.
Teachers spend less than half the time they are supposed to be teaching actually doing so
How can we solve this?
Performance pay for teachers - so incentivised to put in more effort, improve their quality of teaching for their students to get better grades.
Issues with performance based pay (3)
Risk of crowdout of intrinsic motivation
Multi-task problem
Elite bias - neglect weak kids, or make sure they drop out (since bad results reflect badly on teacher)
First issue of performance based pay: risk of crowdout of intrinsic motivation
No longer do it because they want their students to do well, but for other incentives since it becomes a transaction.
E.g wont do it for £1 because feels like its a waste of time, compared to doing it for free since they have intrinsic motivation so feels better doing for free! So have to pay high or not at all
Multi-task problem - teaching to the test
Neglect other tasks to focus on the task the incentive is based on.
Teacher performance pay experiment - Muralidharan - how was intrinsic motivation crowdout, and elite bias eliminated, and multi-task problem (teaching to the test?)
Provide bonus payments based on AVERAGE improvement of test scores.
This eliminated the elite bias, they were incentivised to focus on all students not just the top ones, since drop outs who didn’t take test were assigned low scores.
Eliminated intrinsic motivation crowdout by being framed in terms of recognition for outstanding teaching opposed to accountability
Avoided teaching to test by test design requiring deeper understanding
Findings
No teaching to test - test scores still improved for non-incentive subjects (so teachers increased more effort for all subjects)
Also looked at
performance pay in group bonuses
Block grants to school (money for inputs)
Contract teachers (assistants, lower staff-student ratio)
Feedback
What did they find
Pure incentives (individual and group bonuses are most effective)
Then contract teachers
Then pure inputs are least effective (block grants and feedback(
So individual teacher and group (school) bonuses on performance based pay were the most effective.
Which one more
Individual teacher PP more effective (as shown in original experiment) because group bonuses get diluted (split between school teachers overall
