Height and Entrepreneurship

Academics and think tanks often look at factors that might explain why some countries have more entrepreneurial activity than others. Collectively this group of researchers has argued that immigration, tax policy, the size of the public sector, economic freedom, and a host of other factors are responsible for national differences in entrepreneurship rates.

Many, though not all, of the conclusions of these reports are drawn from examination of the correlation (a correlation of 1.00 means that two numbers move together in an identical pattern) between these factors and some measure of the per capita rate of entrepreneurial activity, such as the self-employment rate or the Global Entrepreneurship Monitor’s total entrepreneurship activity index.

Unfortunately, drawing these conclusions from correlations is a dangerous exercise, given some of the factors that correlate quite highly with national differences in rates of entrepreneurship. For instance, I took a look at data from the Organisation for Economic Cooperation and Development’s (OECD) social indicators to see what is correlated with differences in the self-employment rate across the OECD countries. One of the characteristics most highly correlated with the self-employment rate is the average height of the population, which correlates -0.62 for men and -0.64 for women. Stated differently, countries with taller people tend to have a smaller share of their population that is self-employed.

But does that mean we need to chop off people’s legs if we want more entrepreneurs? Of course not. This correlation doesn’t mean that being shorter causes people to become entrepreneurs. It probably means that something else is correlated with both the height of the citizenry and the self-employment rate. Given the tendency of people from the Nordic countries to be tall, my guess is that the something affecting the data pattern is similar political and economic systems in Nordic countries and the differences between those systems and the ones that tend to prevail in Southern Europe.

In case you’re not convinced that there are a lot of spurious correlations between country characteristics and self-employment rates, consider three others. The correlation between the self-employment rate and the rate at which children under 15 report being bullied is 0.58 for boys and 0.54 for girls, while the correlation between the self-employment rate and the divorce rate is -0.61 and the correlation between the self-employment rate and the share of children under three in childcare is -0.49.

Again, this doesn’t mean that we can get a higher rate of self-employment by introducing bullies into elementary schools, outlawing divorce, or banning childcare. These factors aren’t the cause of high self-employment, they just happen to be correlated with it.

Given the set of non-causal correlations between a country’s characteristics and its self-employment rate, you might want to read more critically any article or report that purports to explain how to make countries more entrepreneurial based on correlations between the countries’ characteristics and their rates of entrepreneurship. To identify the factors that will cause rates of entrepreneurship to increase, we need more serious and careful analysis than just these simple correlations.

* The writer can be reached at