Because someone talks out of their ass and has a lot of money, you believe them based only on anecdote?
Causal effect of education on earnings (using controls, so not just mere correlation) using OLS estimates.
Angrist and Krueger (1991) find estimated effect of between 5%-7% increase in annual earnings for each additional year of completed schooling
Staiger and Stock (1997): 5%-6%
Kane and Rouse (1993): 6%-8%
Card (1995b): 7%
Conneely and Uusitalo (1997): 8%
Maluccio (1997): 6%-7%
Harmon and Walker (1995): 6%
Twin studies:
Ashenfeiter and Rouse: 7%-8%
Rouse (1997): 7%-8%
Miller et al. (1995): 3%-5%
Behrman et al. (1994): 3%-6%
Isacsson (1997): 3%-5%
Another type of estimate, using instrumental variables (IVs), comes up with even higher estimates than these by a couple of percentage points.
After analyzing the accuracy of all these studies, the author that summarizes the findings comes to the following consulsions:
5. Conclusions
Taken as a whole, I believe that the recent literature on the returns to education points to five key conclusions:
1.
Consistent with the summary of the literature from the 1960s and 1970s by Griliches, 1977, Griliches, 1979 the average (or average marginal) return to education in a given population is not much below the estimate that emerges from a simple cross-sectional regression of earnings on education. The “best available” evidence from the latest studies of identical twins suggests a small upward bias (on the order of 10%) in the simple OLS estimates.
2.
Estimates of the return to schooling based on comparisons of brothers or fraternal twins contain some positive ability bias, but less than the corresponding OLS estimates. Ability differences appear to exert relatively less influence on within-family schooling differences than on between-family differences.
3.
IV estimates of the return to education based on family background are systematically higher than corresponding OLS estimates and probably contain a bigger upward ability bias than the OLS estimates.
4.
Returns to education vary across the population with such observable factors as school quality and parental education.
5.
IV estimates of the return to education based on interventions in the school system tend to be 20% or more above the corresponding OLS estimates. While there are several competing explanations for this finding, one plausible hypothesis is that the marginal returns to schooling for certain subgroups of the population – particularly those subgroups whose schooling decisions are most affected by structural innovations in the schooling system – are somewhat higher than the average marginal returns to education in the population as a whole.
https://www.sciencedirect.com/science/article/pii/S1573446399030114
In other words, there does not seem to be much systematic bias or inaccuracies in these estimates.