February 25, 2006

A palpable lie

In response to criticism of a recent special, ABC employee John Stossel has written a column, Myth: Schools Need More Money (or something similar depending on your local headline writer), with the following baloney:

The truth is, public schools are rolling in money. If you divide the U.S. Department of Education's figure for total spending on K-12 education by the department's count of K-12 students, it works out to about $10,000 per student.

The latest official per-pupil K-12 expenditures data shows $7,727 per fall enrollee current spending, on average (for 2001-02). I guess overestimating by almost 30% is Stossel's idea of accuracy. (It's 29.4% if you want to be exact, though it's only 21.1% if it's per-pupil counted as average daily attendance. But since teachers and others have to be hired for capacity, and you don't want kids to be sick or otherwise absent, it's better to look at per-pupil spending gauged against enrollment. Even if you include every drop of spending by K-12 agencies, even if not for K-12 students, and apportion it by enrollment, Stossel is still exaggerating by 11.2%.) And, yes, if one looks at average per-pupil expenditures in constant dollars, they've gone up consistently for most of the century (with a few exceptions: the early 30s, the early 40s, the late 70s, and the early 90s).

But that's not the whole story.


First, average expenditures hide considerable variation, both between states and within states. Some of that between-states variation is accounted for by salary differences from cost of living variation as well as differences in unionization rights and rates. Then there's the within-state variations, in some cases considerable (e.g., Illinois) and in other places dampened by equalization and adequacy court cases (California, New Jersey, Texas, and now New York, among other places, though New York's court opinion hasn't been implemented) or by statute (Florida).

Second, to say that education spending has increased ignores variations in those increases. Let's look at the per-pupil K-12 spending increases by decade (and you'll have to scroll down a bit):









Annualized per-pupil K-12 expenditure increases, 1919-20—1999-2000, U.S.
DecadeAverage per-pupil
spending increases (%)
1919-20—1929-306.9
1929-30—1939-402.6
1939-40—1949-503.7
1949-50—1959-604.1
1959-60—1969-705.1
1969-70—1979-803.0
1979-80—1989-903.1
1989-90—1999-20001.1

(Average increases calculated as the log of the ratio of current expenditures per fall enrollee in 2001 dollars divided by 10. Education Department data starts with 1919-20 school year.)

The 90s data is skewed by six years (1989-90—1995-96) when spending increases averaged 0.1%; the following six years (through 2001-02) had 2.7% average increases. And one needs to be very cautious about interpreting time series when methods of data collection change every once in a while. Nonetheless, even counting just the higher increases in the period since 1989-90, the 90s and early 21st century had national per-pupil increases that were lower than common expectations—and, depending on how much the data-collection assumptions changed, possibly lower than any time since the Great Depression. That fact does not change when one looks at annual data, with the exception of a dip in funding increases in the late 70s and early 80s.

Now, there is considerable debate about how to spend that money, the extent to which the increasing expenses in the 1970s and 1980s were a consequence of commitments to educating students with disabilities, and the distribution question. But simplistic diatribes such as Stossel's deserve to be ridiculed, especially when supported by patently false factual claims. I fully expect Gerald Bracey to do a better job than I in some future writing on Stossel, though I vaguely remember Stossel's getting a Rotten Apple award in the past (or maybe two). Perhaps it's time for a Lifetime Achievement(?) Award.

Listen to this article
Posted in Education policy on February 25, 2006 6:06 PM |