Thursday, March 03, 2011
Wrong as usual
Capo Bernanke is quoted as defending education spending from Repooflican "spending cuts" ... which is pointless in the first place because we know that there won't be any cuts in anything.
Capo Bennie said that economically valuable programs such as education, road construction and research should be sheltered from cuts. [Can't find a direct quote online.]
Along the same lines, a brand-D senator said:
"When they kill education, when they kill research, when they kill infrastructure, they're killing American jobs." [Ditto with direct quote online.]
This hypothesis is correct about infrastructure, dubious about research, and dead wrong about education.
We already know, from every possible set of data, that the correlation between ed funding and ed results is negative. Increasing ed money does nothing for the quality of teaching; it only increases the number and salary of administrators, who make life harder for teachers.
But how about the correlation between ed money and employment?
Professor Polistra tries to answer this.
The solid blobs in this chart represent ed spending as a proportion of GDP, from 1939 to present, derived from a useful set of data at UsGovernmentSpending.com.
Ed spending as prop of GDP tells us when the states (green) and the feds (blue) raised or dropped spending intentionally, not merely when they kept up with inflation.
The shorter blocky blue line represents "civilian employment participation", from BLS. It starts at 1948.
Prof P has drawn a rectangle around the biggest change in participation, the years from 1970 to 1990 when feminists and financiers conspired to force women into the work force, in order to remove a man's ability to provide for his family. This change is so large, and its cause so well-known, that there's no point in trying to compare ed spending in those years.
After that, the last 20 years show a readable pattern, and it's a slight negative correlation. Less spending = more employment. More spending = less employment. Cause? Hard to tell, but the weak correlation is there.
Leading effects also fail. You could assume that ed spending "pays off" 10-20 years later when the "better-educated" students graduate. Nothing in the earlier patterns seems to lead the later pattern.
Thus the Bernanke theory doesn't work. When you spend more on education, you certainly don't get more employment (except possibly among school administrators); when you spend less, you don't kill employment (except possibly among teachers, since administrators are never cut).
Capo Bennie said that economically valuable programs such as education, road construction and research should be sheltered from cuts. [Can't find a direct quote online.]
Along the same lines, a brand-D senator said:
"When they kill education, when they kill research, when they kill infrastructure, they're killing American jobs." [Ditto with direct quote online.]
This hypothesis is correct about infrastructure, dubious about research, and dead wrong about education.
We already know, from every possible set of data, that the correlation between ed funding and ed results is negative. Increasing ed money does nothing for the quality of teaching; it only increases the number and salary of administrators, who make life harder for teachers.
But how about the correlation between ed money and employment?
Professor Polistra tries to answer this.
The solid blobs in this chart represent ed spending as a proportion of GDP, from 1939 to present, derived from a useful set of data at UsGovernmentSpending.com.
Ed spending as prop of GDP tells us when the states (green) and the feds (blue) raised or dropped spending intentionally, not merely when they kept up with inflation.
The shorter blocky blue line represents "civilian employment participation", from BLS. It starts at 1948.
Prof P has drawn a rectangle around the biggest change in participation, the years from 1970 to 1990 when feminists and financiers conspired to force women into the work force, in order to remove a man's ability to provide for his family. This change is so large, and its cause so well-known, that there's no point in trying to compare ed spending in those years.
After that, the last 20 years show a readable pattern, and it's a slight negative correlation. Less spending = more employment. More spending = less employment. Cause? Hard to tell, but the weak correlation is there.
Leading effects also fail. You could assume that ed spending "pays off" 10-20 years later when the "better-educated" students graduate. Nothing in the earlier patterns seems to lead the later pattern.
Thus the Bernanke theory doesn't work. When you spend more on education, you certainly don't get more employment (except possibly among school administrators); when you spend less, you don't kill employment (except possibly among teachers, since administrators are never cut).
Labels: Experiential education
¶ 7:03 AM
