Providing girls one extra year of education beyond the average boosts their eventual wages by 10-20 percent, compared to 5-15 percent for boys. When it comes to such statistical comparisons, there are always a few helpful caveats to keep in mind, especially if you’re about to get showered with numbers as part of week-long festivities marking International Women’s Day. Since many of these comparisons reflect some form of regression analysis, they are vulnerable to a world of pitfalls. For social science statistics, some of the biggest caveats to keep in mind involve independence and noise.
Independence is a technical way of stating the chicken-and-egg question: which came first? Although education can provide access to better paying jobs, the presence of better paying jobs may increase demand for education. Women might not feel compelled to seek more education until after a country’s per capita income increases a little, creating demand for better goods and new services that require more qualified managers and workers. When influence can run both ways, the variables violate the core linear regression assumption called independence. You have to be very careful about how to pick your words and data points to make a proper statement about the affect of more education for women on economic outcomes.
Timing is key, and as luck would have it, there is an organization obsessed with statistics and how to effectively use them, that provides online tools to help examine time and regressions. Led by Hans Rosling, Gapminder’s online tools present time-series regression analysis using various international data sources. Using Gapminder, you can see for example how global data generally supports the story that achieving gender equality in education generally comes before raising per capita incomes (move the slider manually to get a better look).
When you watch how countries move through the comparison over time, notice that approaching gender equality takes a lot of time, and per capita income doesn’t really start to rise until very late into that process. While that’s good news for those countries moving away from gender inequality in education, this lag does two things: it shows how the relationship clearly isn’t quite linear, because a change in one variable doesn’t correspond to a constant rate of change in the other; and it brings us to the second caveat when it comes to statistics, noise.
You may already be asking yourself, how much can happen in the lag between gender equality in education and when incomes actually start to rise? A lot, of course. While the chart shows it’s clear that higher girl/boy ratios occur first, there are many things that could happen in the meantime to raise or stagnate per capita income. Governments can change, institutions can fall apart, institutions can emerge, entire economies can re-organize.
Gender parity in education and income per capita are linked to a number of other co-related variables (also called confounding variables), which dilute the poignancy of making a statement with them. Each may be more strongly related to one of these other variables. The strengths of each relationship may change over time. There’s just so much going on, so much noise, it’s difficult to get a firm grasp of how much this really affects that, at the moment, and vice versa.
It is possible to cut out some of the noise, or at least find relationships that could only get help from what can’t be accounted for at the moment. In today’s featured fact, we identify just four groups: girls with just the average number of years in school and girls with one more year, alongside boys in the same two groups. Then we’re seeing what happens to them later, in terms of their future average wages. Although there are many things that could account for how much more each plus-one year cohort makes compared to the average for their respective genders, there are also vast untouched vestiges of patriarchal power littered across the world, making it all the more notable that one more year of schooling beyond average moves the wage distribution for girls farther than it does for boys.
In other words, in the countries studied, girls get more out of education than boys. There’s a lot of noise around what might make them finish out their secondary or higher education. There’s a lot of noise around what more educated women do to inequality or income per capita. But cut it down to just one cross-country cohort of girls with one more year of schooling than average, who have very little falling in their favor besides education, and they are still getting more future wages out of that extra year than a cross-country cohort of boys.
There’s no telling exactly what advantages the extra year brings to the girls. It could be additional skills, additional personal contacts, additional confidence, or most likely some combination of these and other things. CIPE’s Tashabos program addresses both entrepreneurial skills and skills to organize and voice the needs of one’s community of fellow entrepreneurs. A slim majority of more than 23,000 students in Tashabos classes are female. It will be very interesting to see what their eventual wages will be.
Published Date: March 08, 2010