Two weeks ago the OECD published its education report for 2008. The report got the usual media coverage, i.e. – an extremely superficial and sloppy coverage. I took some time to run over the data, as I do every year, and added it to my notes on previous reports, and on research done by Dan Ben-David from the Tel Aviv University. Here are some of the mantras that we usually hear, and the way it relates to the data:
1. "We spend too little on education. We should spend more": First, how much is enough? Furthermore, if two countries spend the same amount but the number of children in the first country is half than the number of children in the second one, is it reasonable to say that their investments are the same? Obviously not! Thus this assertion is not well defined.
2. "We spend too little on education per student. Our spending per student are the lowest amongst the OECD": The OECD tables does show it. But, alas, the OECD tables show expenditure in terms of PPP. This is OK if we want to send our children to study in Norway; but as our children study here, we should normalize the expenditure (in terms of PPP) by the standard of living, as offered by Dan Ben-David. Following this line of though leads to the conclusion that we spend much more than many of the OECD nations. Summarizing Ben-David's data (table 6) into a chart reveals that the more nations spend the less their students tend to achieve in the PISA exams:
The chart thus tells quite a different story than the assertion above. One possible interpretation of the data is – bad education costs more!
3. "We have bad ignorant teachers and bad teacher training. Such teachers cannot lead our children to achievements": This is very true (as shown by numerous reports along the years). But is it relevant? Does the quality of teachers influence the quality of education?
A good proxy to the quality of teacher could be the ratio "teacher's wage: living standards". We shall thus use the ratio "teacher's wage: GDP". If we measure achievements by the students' scores in PISA then we get a correlation of 0.64 if the wage is the average wage of a senior teacher in primary schools and a correlation of 0.82 if the wage is the average wage of a senior teacher in high school (p<0.05). Thus the ratio "teacher's wage: GDP" can explain up to 68% of the variance! How should we interoperate that? Well, teaching is a career. Everyone who considers teaching as an option will ask "what will be my future wage in this line of trade?". If the wage, compared with the standard of living, is low then many qualified young men and women will not choose teaching as their profession. As the ratio above increases many more talented people will join teaching at schools. This will have an influence on the entire system, as senior high school teachers usually teach lower grades first.
4. "Teachers wage is the most important factor influencing achievements. Thus, we should at once increase the education budget". NOT! We had many such increases which led to nothing but a decrease in achievements (see Dan Ben-David's website showing an increase in the expenditure per student from 1992-2002 and a drop in achievements in those years). The system is sick, and each increase in the budget will be wasted by the endless bureaucracy. The system should be change to make it possible for teachers and students enjoy the budget in full, and make achievements worth the spending!
5. "Our classrooms are too crowded. Making classrooms smaller will make things better": OECD data shows a correlation of -0.3 between class size and achievement (p<0.05) thus classroom size contributes at most 10% to the variance. As many of the leading countries on the achievement side have extremely crowded classroom (up to 46 on average) it is very hard to deduce that making classrooms smaller will do any good. Furthermore, making classrooms smaller and nothing more can make things worse; Making classrooms smaller means more classrooms are needed. More classrooms means more teachers which means one (or two…) of two: a. Letting teachers with a part time position teach more or b. Hiring new teachers.
Option (a) is bad – disregarding a minority of idealists, people who are willing to hold a partial position and earn peanuts are not the people we want as teachers…Option (b) is worse – this means digging under the barrel…
6. "Decentralization/Privatization means that rich kids will have education and poor kids won't as opposed to the current condition": Data shows that it is the current systems that differentiate the rich from the poor. It shows that we have the highest proportion of 9th graders who study less than 2 hours of math or less then 2 hours of science a week at school amongst OECD countries. At the same time we have the highest proportion of 9th graders who study more than 4 hours of math or more then 4 hours of science a week in private lessons. Thus the current system leads to an immense differentiation in knowledge between rich and poor, as was shown by the PISA exams.