Social capital broadly speaking refers to the degree to which people in a community tend to trust and cooperate with each other. It is a concept that is obviously country-specific, and has been shown to affect economic development.
In general, it is thought that social capital is hard to change. It is usually passed down via one’s family. Algan, Cahuc and Shleifer (2013), however, hypothesize that perhaps education, and more specifically not the quantity but the method of education used in schools can help build social capital.
Education in general is a double-edged sword. The reason for this is that education is more than just teaching a certain pre-determined curriculum and skills to a bunch of students. Education can also have, often unintended, effects on people that they may not even realize consciously.
Of course, a government or school can change things about education (such as the curriculum) that change the “hard” results of education, that is the amount and type of material and skills a student takes away from school. However, educational decision-makers can also affect other outcomes, for instance the attitude of students towards certain things. In a previous post I wrote about how one’s personality might be affected by different educational systems. In this post, I will focus on how students’ trust towards each other and their willingness to cooperate (which can further translate into more important things later in life) are affected by educational decisions. These two examples show that one has to make very careful decisions when it comes to education because – as in most developed countries students spend at least 16 years in school – education has a huge effect in molding the minds of the population.
When it comes to social capital, the authors of the paper in question posit that in countries/schools where teachers teach using lecturing and a sort-of one-way communication (vertical teaching) as opposed to letting students work and discuss with each other (horizontal teaching), students will be less trusting of others. These countries will subsequently have lower social capital, which will translate for instance into a lower trust in institutions, public servants and politicians; and a lower participation in civil life such as a lower membership in associations.
Cross-country differences between lecturing vs. group work are shown in the figure below.
In a cross-country setting the authors find that more vertical teaching leads to lower general trust, lower association membership, lower trust in civil servants, and a more frequent perception that civil servants are unfair. These results are significant at the 1% level. Control variables include school expenditure, income per capita, and average years of education; none of which are significant consistently for all these measures of trust mentioned above.
Using micro data, the authors find a similar trend. Specifically, students who work more in groups are more inclined to believe in cooperation among students, and between students and teachers. Students who are taught more often with a vertical method are less inclined to do so. Control variables that are consistently significant include being female (+), number of books at home (+), and whether the teacher’s goal is to promote cooperation (+), where a + means a positive relationship between the variable in question and the belief in cooperation. Furthermore, class size appears to be negatively related to a belief in cooperation, albeit only at the 10% level.
With the same micro data, if students work in groups they’re more likely to have an association membership, trust institutions, participate in political, and in social life.
But just how powerful are schools in scultping those young minds? Quite interestingly, the results might be driven by students who receive less attention at home. Specifically, the authors divide the sample into schools whose share of students from a low socioeconomic background (i.e. below average) is above the average share of students from a low socioeconomic background in the country (from now on: poor schools), and into rich schools (i.e. below average number of low SES students). Then they find that while in poor schools group work/lecturing is indeed a significant predictor of belief in cooperation, in rich schools the effect is insignificant. This means that it appears that low SES students might be more receptive to the educational system’s inputs. In my opinion, this may potentially be because these students receive less input (i.e. attention, parenting, etc.) at home. So they go to school with more of a clean slate; their minds are not yet made up on issues like trust.
The results could be driven by reverse causality, i.e. that it is not more group work in school that causes higher trust, but higher trust causes more group work in school. For instance, if a community has higher trust, it might impose horizontal teaching practices on its schools (whether consciously or unconsciously), or teachers specializing in horizontal teaching may be selected or are self-selected in these communities. To tackle this issue, the authors look at within-school variation in teaching practices. This helps because then effectively, they control for the trust level of the community in which the school operates. The results are confirmed by this analysis, indicating that they are not driven by reverse causality. I.e. it is indeed group work in school that causes higher levels of trust and not vice versa.
Two additional robustness checks include confirming that horizontal teachers are not better teachers (in general, teaching style is not correlated with teaching quality), and that horizontal teachers are not nicer teachers. Results are thus not driven by teacher quality or kindness.
So what do we learn from all this? It seems to be the case that horizontal teaching practices (e.g. group work and discussion among students) help build trust not only among students but in general. Trust can lead to a higher involvement in civil and political life in general. It has also been linked to economic development (I’m not sure about the causality here though). Does this mean that countries should adopt more horizontal teaching? Perhaps, but not necessarily. As can be seen in the second figure above, there are notable exceptions to the rule (such as Ireland and Japan where high trust is coupled with vertical teaching). Horizontal teaching is not the only way to increase trust, but it might be something worth considering for some countries. Countries highly focused on vertical teaching could introduce the horizontal teaching practice in some of their less important school subjects to expose students to this method and increase social capital gradually.
A final thing that comes to my mind is how these attitudes that are molded by education (personality or trust) can affect the population’s long-term political leanings. I think the potential to sway political positions is huge. Therefore, when a government proposes seemingly harmless or neutral educational reforms, it must be subject to extensive scrutiny. For instance, in 19th century Italy and France local languages, cultures and identities have effectively been exterminated to give rise to the Italian and French cultures, languages and national identities.