Does Money Matter (Part 2)

A few days ago I wrote about a Texas Monthly article from 1979. Well this one is older. Philip send me this article from 1966. Written by Christopher Jencks. Those who know him call him Sandy. I do not know him.

But he's still an active Harvard faculty member today; advisor to my compadre Jal. He's a lion in the field.

44 years ago, Jencks wrote:

As a rule of thumb, America spends about half as much educating the children of the poor as the children of the rich. The difference derives from two factors.

First, the annual expenditure per pupil in a prosperous suburb is usually at least fifty percent more than in a slum in the same metropolitan area.

Second, this additional expenditure, in combination with better family and neighborhood conditions, encourages suburban children to stay in school half again as long as slum children (from kindergarten through college, instead of from first through tenth or eleventh grade).

If America were to try to provide all her children with equal opportunity to develop their talents, obtain ample adult incomes, and share in controlling their own and their community's future, this pattern of expenditure would probably have to be reversed. If we wanted to offset the misedueation which takes place in a slum home and neighborhood, we would probably have to spend twice as much on formal education in the slums as we do in the suburbs.

Instead of spending less - often much less - than $500 per child per year for education in the slums, we would have to spend more like $1500 per year.

But of course this is not going to happen. Almost nobody really wants to make America an egalitarian society. Ours is a competitive society, in which some people do extremely well and others do equally badly, and most people are willing to keep it that way.

(Interesting to recall the term "slum." Now we say "inner-city." Or "urban.")

Jencks hopes for $1,500 per year in 1966 dollars for the education of slum children. He is sure it will not happen. How did his prediction hold up?

What is $1,500 per year in 2010 dollars? It's $10,034.

Hmm. That is about what we spend per year on slum children today!

So at least that aspect of what Jencks thought would never happen...has actually happened.

Now how does that compare to suburban districts today? It varies by state. According to the Ed Trust, high-poverty districts in Ohio get 26% more than non-high-poverty ones; in Michigan, about the same; in Illinois, high-poverty districts get 6% less.

When added together, Ed Trust in 2009 said poor districts get about $773 less per kid per year from state and local sources. But the feds step in with about $1,500 for poor districts.

Net? Poor districts spend $700 or so more per kid these days than in non-poor suburbs.

However, another aspect of Jencks' vision remains unrealized. Remember, he argued that spending for the slum kids needed to be double the suburbs, not about the same, not simply equalized.

Has the spending helped slum kids?

The leading expert on this question is probably Stanford economist Eric Hanushek. He's has long argued that there does not seem to be any correlation between higher resources and increased achievement, not just in USA but around the world. Not unless, he says, you make major changes to incentives.

Using the same test we applied to baseball, football, and basketball....we'd have to say that K-12 is like football, where spending more doesn't correlate with different results.

Now don't get totally depressed. On the other hand, it does seem like slum kids are better off than in 1966. For a while, the racial achievement gap declined, though that slowed/stopped the last 2 decades. Also, kids of all races do better now than back in 1966.

Look at NAEP data here. Click on the different tabs. This is just one example, that of 13-year-olds.

Of course, hard to pinpoint the cause of these NAEP gains.

Might not even have anything to do with schools themselves. Changes in parenting? Technology? Safety? Less racism and a vastly expanded black and Hispanic middle class? (If Jencks thought it impossible that we'd equalize slum and suburban per-student spending, I wonder what he would have thought of the notion of a black president).

So here's where we are.

Spending more money does not, at the macro level, correlate with kids learning more. In "slums" that do spend the most, even more than in suburbs, like Camden NJ and Washington DC, we still get the same incredibly low outcomes, even lower than other slums.

We used to spend a lot less on a "slum" kid than a middle-class kid from a nearby suburb. Now we spend about the same. Jencks thought we should spend a lot more on a slum kid. We're not there yet.