Nova workboard

a blog from young economists at Nova SBE

The number of students per teacher in OECD

1 Comment

“Education is the most powerful weapon which you can use to change the world”

Nelson Mandela


Today I will write about some interesting data that I found on Pordata, related to the topic of education. I will analyze the average number of students per teacher in lower secondary education (ISCED 2).

First we need to understand education as a cumulative process and, due to that, in principal, the steps that one individual take in their education has impact in their future life.  The point is to understand how correlated is the impact of the institutions quality with the indicator that I presented. As I don´t present an econometric or an empirical study about that, I will use this indicator to see, empirical, what is the relation of that and the kind of the country presented. Notice that, I only present information related to the European countries.


According to the graph, for the countries presented, we verify that, on average, the number of students per teacher, in general, decreases from 2000 to 2011. Why does this happened? One obvious reason is because the birth decreases, so the number of students per teacher decreases because we have less children in European countries. According to this logic we can argue that, if this happens, probably the quality of institutions in education was not the reason for this change. But, why I am talking about the quality of institutions? What I am trying to focus is that, in a very simple model, the decision of how many students each teacher can have in each country is a decision for the government that can represents and determine the quality of education on these countries, which is a different indicator from class size and it is not a good predictor of educational outcomes. Due to the fact that in primary school learning is much more important than signaling, the quality of institution assumes a very important key, in particular, teachers assumes an important role in children. For the secondary education maybe it is less important. The decision of how many students each class will have depends on the number of teachers that one country has in certain areas and, for that, is natural that these values varies between countries. However, according to the figure we can see that European countries are homogenous in this indicator. Afterwards, what we are interested to see, as economists, is that how much will be the return on growth in each country for investing (or not) on education.

Concluding, the main point is that few students per teacher in first years of education assumes an important key in student development. In last decade, we assisted to a decrease on the indicator presented, as a general trend is most of OECD countries, but the reason of that can be associated with a demographic change and not a change on quality of institutions of the education sector. Thus, this indicator is not a good measure of the quality of institutions.

Link to Pordata:

Pedro Luís Silva

Research master in Economic, #87


Author: studentnovasbe

Master student in Nova Sbe

One thought on “The number of students per teacher in OECD

  1. The main focus of Pedro is on the analysis of the students per teacher ratio in lower secondary education. He invokes how the quality of institutions can be correlated with the mentioned indicator, arguing that it might have not been the main cause of the evolution of the indicator from 2000 to 2011.

    He defends that from 2000 to 2011 the average number of students per teacher declined for the majority of the countries. One reason he invokes for this scenario is the decline in birth rates across these countries. Is that a plausible reason? He does not show any statistics regarding these rates, however, if we look at the data by OECD, for Spain the fertility rate went from 1.23 per cent to 1.38 per cent in 2010. In Sweden, it has also increased from 1.55 per cent to 1.98 per cent for the same period. Probably, the relevant change in fertility rates was in the period 1990-2000. It means that for some countries it is not a good argument to infer about the decline in the average number of students per teacher based in some demographic indicators such as fertility rates. Nevertheless, it is true that some countries presented in the graph have experienced a decline in birth rates, which can explain in part the dynamics of the indicator.

    Regarding the quality of institutions, Pedro is not clear at all: according to him, the quality of institutions is a decision from the government. This is true in part: the government is on charge of investment in infra-structures in public schools, for example. However, the decision concerning the allocation of students and teachers is often in charge of the headmaster or school-director, with the exception of some developing countries, where the students per teacher ratio is high and there are no defined rules at all (this is the point that lacks in the argument of Pedro). Furthermore, from my view point, there is no distinction in the text among quality of institutions and quality of education. These are two different concepts. The former stands for the quality of the resources that schools provide (libraries, study rooms, whiteboards); the latter is connected with educational outcomes (that can be measured through scores in tests by students, in general, school achievement). But, obviously, we should not totally disregard a link between the two factors.

    In order to assess the evolution of the indicator, it seems reasonable to look at other factors that might have exerted a stronger influence, besides fertility rates. One should look at educational policies implemented by the government (regarding admission of students and teachers, for example). These have changed considerably in Portugal for example, where teachers have to do admission exams to be eligible to teach.
    As Glewwe, HanusheK, Humpage and Ravina (2011) pointed out, the student per teacher ratio did not seem to have a significant impact on enrollment or learning. This is not a good measure of the quality of education.


    OECD Statistics
    Lectures slides

    Nuno Lourenço, 85