Correction there IS correlation between per student school spending and test scores

Unadjusted for demographics, spending has no impact on test scores.  Hat Tip  Super-Economy - a good Kurdish/Swedish/American econ blog.


However when you focus just on non-hispanic whites, the relationship is strong.

Likewise when you look at just blacks, spending per pupil correlates highly.

So does this mean more money means better outcomes?  I think not.  First of all, more money is a proxy for higher incomes.  New Jersey, an affluent suburban state, has much more highly educated and compensated parents than Mississippi.  In addition, the cost of living is much higher - probably twice as high in NJ as in MS.  Adjust for those factors and the correlation will weaken, though not disappear.  The real question is this:  if you equalized MS spending up at NJ levels or cut NJ down to MS levels, would the gap narrow appreciably.  I think not.

The other brutal fact is that both MS and NJ (and the rest of our nation) got the same results for one third the money (in real dollar terms) 40 years ago.  We've spent a fortune and have nothing to show for it....but a bunch of rich unions.