One thing that should jump out immediately from Figure 1 is that this is a really heterogenous group of countries with utterly dissimilar higher education systems. You have two countries with zero tuition (Sweden and Finland) near the top; you also have the US and the UK where average tuition is in five figures as well as Tanzania, where tuition is highest of all, at least relative to GDP/capita (a year at the University of Dar Es Salaam costs an amount equal to about 60% of per-capita GDP, or four times what it is here in Canada). And then there are countries in the middle like Canada, New Zealand, the Netherlands and Taiwan.  There are countries where private universities are the dominant form of higher education (Japan, Taiwan), places where it is not predominant but still quite common (US, Tanzania), to places where it is non-existent (Netherlands, Sweden). Overall, it’s hard to say that one type of fee-regime or ownership model causes higher levels of loan use. Different countries with different structures use loans for different policy purposes.

In addition to looking at how many people get loans, it’s worth looking at how much those loans are worth. In Figure 2, I standardize this across countries by dividing the average award amount by GDP per capita. In most countries, the size of the average annual award tops out at 20% of GDP/capita and is sometimes significantly lower (in Canada it is just 12%). The two outliers are the United Kingdom, where every student is eligible for a stonkingly huge loan to cover fees, and Tanzania, where the average loan is equal to over 120% of GDP (which mostly tells you that the income gap between the educated elite and the rest of the population in Tanzania is pretty large).