10/13/2009 Lecture 10 The MDC goals call for universal primary education by 2015, and equality between boys and girls at all levels. We've moved toward this goal impressively. From 1999-2000, enrollment in sub-saharan africa went from 54% to 70%. Anne Case (The Primacy of Education) studies the correlated benefits to individual of education - health - fertility - future wages/income rise Schultz: human capital is benefit to onesself---win nobel prize for this. Being educated allows you to better understand new technologies, new health ideas (germ theory), gain control over ones life (fertility, life expectancy, etc.) Case also mentions that it's hard to tell whether being rich leads to more education, or whether education is what gets you there. Returns to education may be lower when there is nothing to do/no technologies to use in the country you are in when you graduate. -> Example: green revolution, india--- In last 20 years, there has been an increase in "skills-biased technological progress," which has disproportionately improved income of those with education. Benefits of education to society - Children who are educated may educate their children - Educated people who avoid health concerns will not infect others - Foreign investment might be more desirable in a more educated country - Might make it easy to start new industries - Neighbors might pick up skills or immitate techniques from educated folks - More opportunity for uneducated people to have managers where such a relationship is complementary Costs to me if others around me are educated - Too many educated people will crowd each other out of the market - Lowers the wage of other educated people if job supply doesn't match new demand - Parents may lose current income at expense of future income - Policies from educated folks might benefit other educated folks, leaving the uneducated behind So Easterly article does not necessarily go against this. There are costs and benefits to education in society, and it's unclear what this looks like at the aggregate level. Let's look at the evidence - There is a _huge_ correlation between years of education and GDP - Easterly shows that even though this is the case, there is no correlation between _change_ in years of education and _change_ in GDP at the aggregate level. In general, things get worse for Africa and better for Asia. - Duflo shows graph of education in Indonesia that shows ~7% jump in income for each additional year of primary education. It actually turns out that this trend is robust across the world: it's somewhere between 8-11%/year. - Countries in top decile of education have about 8 more years of education than those at the bottom. They should have a GDP 80% highe if private returns were only part of the story. In fact, they are 15x richer! -> One possibility: there are many social benefits to education, which lead to higher GDP growth. So looking within a country, you only see income gains. But between countries, you see large social externalities benefitting them. -> Another possibility: the countries that see 15x GDP increase have more opportunities/less corruption to use education once the educated people graduate. -> Another possibility: (Easterly) Countries that are more educated may be different. People may choose to get educated when they see good prospects. If you live in a country you think is going somewhere good, you'll invest more in your child's education. Easterly---if education led to more gdp, we would see correlation between change in education and change in gdp. However, there doesn't appear to be such correlation. Problems with Easterly's conclusion - Education is not well-measured at the country level: hard to compare enrollment, attendance, difference in quality. Comparing differences in the output just doubles error (subtracting introduces two error terms), so you are measuring more noise. - In countries where education has increased fastest since 1960, the quality of education has decreased. Rapid education happened by doubling class size, which decreases education quality. - Some countries that had fast education also had real problem (civil wars), which weakened them. Easterly isn't saying education is ineffective. It's probably necessary for growth. But it's not sufficient. There are external factors, namely wars, corruption, and systemic acceptance. He goes on to suggest that the correlation of education to growth is mostly reverse causal: the rich like to educate their kids. This is where Duflo begs to differ, since at the individual level (see the Indonesia data), things do seem to improve with education. Side idea: "Years of education" is misleading! One year of American education has different value than one year of education in an African country. We'd like to measure causal effect of education. Are there reasons why the "counterfactual" wages of someone who received a higher education are higher? intelligence, or ability to pay? It's hard to do a randomized test, since it's hard to randomly assign education. Instead, look at differences in differences study to see how things change within a country at different times. Experiment: 1973 oil shock. Indonesia is oil-producing country, and used the money to build schools. - Educated those age 12 or youner starting 1973. People who were older did not benefit. - 60K schools built in 5 years (more than doubling # schools), in regions that were initially lagging behind in terms of education. Results: (Duflo) compare two cohorts both in 1995. Compare kids who were 12 or older to those who were younger than 12 (and got educated). Compare regions that got many schools and regions that got few. Compare difference between both age groups across places with few and a lot of schools. -> compare number of years of education between young and old as number of schools change. Each extra school per 1000 kids results in .2 extra year of education difference between cohorts. -> compare wages between young and old as number of schools change. Each extra school per 1000 kids results in 1.5% increase in wages. -> paper rules out reasons for increase in wages other than increase in number of years of school. So if school doesn't increase wage other than by increasing years of school, then value of 1 year of education = (1.5%/.2) = 7.5%. This verifies the 7% increase that we saw through the original chart of education affecting wage increases. Someone used the same method to measure effect of years of education on child health. Saw that it does increase health. Also, saw increase in fertility. So at the individual level education helps. Next steps: address Easterly's concern that it doesn't affect us at the global level. Prof. Duflo said that no one has studied the benefit to a country with some technology of education in another country that prepares foreign workers to come and use the technology.