O Canada

According to George Will, "Sen. Daniel Patrick Moynihan once puckishly said that data indicated that the leading determinant of the quality of public schools, measured by standardized tests, was the schools' proximity to Canada. He meant that the geographic correlation was stronger than the correlation between high test scores and high per-pupil expenditures." But as Kevin Carey observes it's highly unlikely that Moynihan meant any such thing -- since the highest-spending states are all near Canada.

Comments

MY:
You realize that Atrios named Will the "Wanker of the Day" over this column. Does Will realize what this means about school in the Republican strong holds?

Posted by Joe Klein's conscience | April 25, 2008 9:00 AM

Tennessee needs to investigate those few pesky colored counties ruining their otherwise clean, white appearance.

Posted by steve duncan | April 25, 2008 9:03 AM

I'd like to see more breakdown of the $17,250 groups. I bet there's counties in Mississippi getting by on what they collect on recycling cans.

Posted by daveNYC | April 25, 2008 9:06 AM

I'm surprised by Wyoming. Many of the higher spending areas have high costs of living, others are thought of as predominantly liberal. What's with Wyoming? Do they spend $9000 per student busing them 200 miles to school each way?

Posted by Njorl | April 25, 2008 9:13 AM

While interesting, the map really needs to be adjusted for population density. The high per pupil spending in Wyoming, eastern Oregon, central Nevada, northern Alaska, etc. has got to be in large part due to the extremely small number of students in those locations.

Posted by Just Dropping By | April 25, 2008 9:13 AM

George Will God hates you.

Posted by freddiemac | April 25, 2008 9:18 AM

Maybe proximity to Canada is a rough guide on a national level - but at least within New England that rule breaks down pretty quickly. If you move to Northern New Hampshire or anywhere in Maine more than 40 miles from the coast in search of good public schools, well, good luck to you. I'm guessing the same holds true for upstate New York as well.

Posted by vanya | April 25, 2008 9:22 AM

I'm sure this correlation between test scores, spending, and geography has nothing to do with the fact that many northern states invest more in education and social welfare in general (whether it's Northeastern/Pacific Northwest liberalism or Great Plains/upper midwest progressivism) and most southern states (whether it's the Dixiecrat South or the Goldwater southwest) invest less. No, no--it must be Canada's fault.

That map reminds me why I'm so happy not to be teaching in Tennessee anymore.

Posted by Marc | April 25, 2008 9:22 AM

The high spending in Alaska is also related to the fact that there are no roads in northern Alaska, so everything must be flown or barged in, plus most regular supplies cost 25-100% more than down south. Also gas is more expensive.

And even when you have a school, which is usually a couple trailers, you have to find a teacher to teach villagers (200-300 people is a big village), and to get a college educated person to live in the Alaskan bush (-60F winters, no stores or entertainment) with nothing else around them takes money.

I don't know what Wyoming's excuse is, though.

Posted by anon | April 25, 2008 9:23 AM

Given that employee pay is the largest expenditure in education, wouldn't this be a lot more useful if it were adjusted for cost of living expenses? Comparing absolute spending in New York and rural Alabama is kind of weird.

Posted by mf | April 25, 2008 9:24 AM

Lesson learned: if you want your kids to get a good education, Northern Alaska is the place for you.

Posted by MS | April 25, 2008 9:24 AM

It's certainly possible, based on this map, that proximity to Canada correlates with test scores better than per pupil expenditure. Just for instance, I'll bet Iowa has higher test scores than New Mexico, even though New Mexico is darker on this map. And I wouldn't be surprised if Missouri had higher test scores than Georgia. One would have to run actual regressions to answer that question, although obviously calling people "wanker of the day" would be a lot more fun.

Posted by y81 | April 25, 2008 9:27 AM

I'd guess this very same map could be used to reflect rates of cigarette and alcohol consumption and percentages of obesity in adults, inverting it so that light colors reflect high rates of each with darker colors showing lower numbers. Yes, there'd be notable variances but I think it would be mostly accurate.

Posted by steve duncan | April 25, 2008 9:31 AM

Everything is better closer to Canada, and even better in Canada. Hot weather apparently makes people angry, selfish, and stupid.

Posted by Gary Sugar | April 25, 2008 9:33 AM

We hyah in Jo-jah don't need yer dang fancy pants ay-leetiss big gubmit skoolz no way. We got Neel Boortz and our local mega-church who give us all the edumikashin'in we need.

Posted by El Cid | April 25, 2008 9:36 AM

I'd guess this very same map could be used to reflect rates of cigarette and alcohol consumption and percentages of obesity in adults

Also, educational attainment of parents.

Can anyone link to evidence on where public schools are highest quality? I'd find it surprising that, say, West Virginia's are better than Florida's.

Posted by right | April 25, 2008 9:37 AM

Can anyone link to evidence on where public schools are highest quality? I'd find it surprising that, say, West Virginia's are better than Florida's.

Posted by right | April 25, 2008 9:37 AM

The public schools in Florida are notoriously terrible. On the other hand, they give out tons of money to students for college. For example, the requirements to get into UF (the state's flagship university) are higher than the requirements for a full ride on Bright Futures. Thus, nearly every undergrad at UF goes for basically free.

Posted by lfv | April 25, 2008 9:49 AM

So the Detroit Public School system could set up floating schools about 1000 ft into the Detroit River and all its problems would be solved.

Posted by waatmf | April 25, 2008 9:52 AM

I can't imagine it would be that difficult to just do a comprehensive correlation between academic performance and per-student spending either at the county or the district level, and take a look at the results. Computers make lots of projects pretty trivial.

I tend to thing the correlation would be pretty low, at least in comparison with non-financial factors. After all, doesn't Matt's own DC have about almost the highest student spending and also almost the lowest student scores? I'm aware of enough other such major situations to make me extremely skeptical that spending explains much at all.

Consider also that spending rose dramatically during the exact same decades that student performance fell dramatically.

Posted by RKU | April 25, 2008 9:54 AM

It would be interesting to see these numbers adjusted not only for higher wages in the north, but for the costs of heating the schools. I suspect that would pull the performance numbers down enough that the remaining differential would be explained by the normal indicators -- socioeconomic status, educational level of the parents, non-English-speaking kids, and so on.

Posted by veblen | April 25, 2008 9:56 AM

But as Kevin Carey observes it's highly unlikely that Moynihan meant any such thing -- since the highest-spending states are all near Canada.

Huh?

How does this in any way reflect on what Will said? Geography (distance from Canada) may be correlated with spending. But so what? How does that change the fact (if it is indeed a fact) that geography is more correlated with high test scores than spending is correlated with high test scores?

Posted by Al | April 25, 2008 10:02 AM

I think Moynihan's original point was that statistical analysis may not give you an actual answer. Facts are facts; it's interpretation that can be the bitch.

Posted by Ed Callahan | April 25, 2008 10:07 AM

y81-

Did Will run those regressions? No. Did he link to such a study? No. Did he take a quote from 20 years ago and pretend it to be true? Why yes, yes he did. But, oh better not call him a wanker! We must take seriously his completely made up argument because it might be true!

Posted by Rob | April 25, 2008 10:14 AM

Closest to Canada. Farthest away...from Mexico.

Posted by Curmudgeon | April 25, 2008 10:31 AM

How much of the money spent by Kansas was related to trying to push religion (a.k.a. "Intelligent Design") into their public schools?

Posted by LFC | April 25, 2008 10:51 AM

The crux of Will's argument is that for all the money we've spent, our scores are still going down, so that's evidence that it's money going down the drain.

But here's the problem with his analysis - if increased spending in education, higher teacher compensation, and lower class sizes has resulted in falling test scores, then does he believe we should offer lower salaries and increase student to teacher ratios in order to reverse these trends?

Or perhaps, has our spending actually kept scores from falling even farther than they otherwise would have given the societal changes that have come since our "peak" (as Will sees it) in 1964? Never mind that the number of students taking the SAT and going to college have increased dramatically over the last 40 years. The "decline" must be because schools are performing ever more poorly, not that they are succeeding in sending an ever wider cross-section of our society into higher education.

Posted by mbuchel | April 25, 2008 10:56 AM

Al,
It should be noted that based on the Will quote you are not defending what Moynihan said, but rather Will's odd take on what Moynihan said. Given that closeness to Canada is highly correlated with expenditure on school, as shown in the graph below, it is more reasonable to take Moynihan to be using "closer to Canada" as a stand in for higher spending than as a contrast with it.

It seems more likely that it was a sarcastic way of indicating the better performance of the more Liberal north compared with the more conservative southern states.

But it should be clear from the Will quote that the silly part of it is not actually taken from Moynihan, but is read into Moynihan by Will.

The only interesting interpretative question about the quote is what Moynihan meant "closer to Canada" to be a stand in for. It could be, as Ed Callahan suggests that it is a stand in for the arbitrariness of random correlation distributions. But my guess is that (exactly contrary to the way that Will is taking it) it is a stand in for the approach one finds towards education in the Northern states vs the Southern states.

Posted by Lon | April 25, 2008 11:04 AM

As I recall, the correlation between educational spending and test scores depends on what category of educational spending you look at. For example, the correlation between overall spending per pupil and test scores tends not to be statistically significant, but if you back out transportation costs you can get something statistically significant, more so when you also back out everything but operational costs, and even more so when you also back out everything but instructional costs.

Which, of course, makes sense.

Posted by DTM | April 25, 2008 11:09 AM

Will and Moynihan were famously close, but of course those who were in jr high when Moynihan died wouldn't know that. Will has Moynihan right, and the kids have it wrong. The column is one that Moynihan could actually have written. The puzzle then and now is, why are Iowa's scores better than Arkansas's. Will and Moynihan agree on the answer, and it isn't funding.

Posted by Thomas | April 25, 2008 11:17 AM

I would be curious to see this map laid over one indicating property taxes. I grew up outside of Atlanta, and I can promise you that the quality of education there, despite the apparent high cost per-pupil this map shows, is no indicator that the schools are any good. The problem is parents are paying far too high a price, out of pocket, not out of their county property taxes, to pay for their students.

Teachers are still paid ridiculously low salaries in many counties, and so stay for only a short time in some areas, taking their expertise with them. Just because the school's baseball booster club is raking in $$ to pay for a new baseball field, does mean the quality of education, guidance counseling, etc. is top notch.

Posted by CHoward | April 25, 2008 11:28 AM

The leading determinant of school quality as measured by standardized test score is the proximity to... white people. And there are higher percentages of white people in the states that border Canada (and those close by, like Iowa) than those that do not.

The dollars = quality outcomes is simply false. Drill down futher than the state level to see if this is true. Throwing dollars at urban schools in Detroit, Chicago, Milwaukee, Minneapolis, etc. does not correspond to higher achievement. The outcomes based on race are profound, even within the same school district:

http://www.wpri.org/Reports/Volume%2020/Vol20No6/Vo20No6p1.html

The higher dollars per student spent in the northern states is a function of stronger teachers unions and a greater willingness/ability of the citizens to endure higher local taxation.

The quality of the product depends on many other factors, money being somewhere down the list from parental involvement and parental giving-a-damn. Without that, the money just does not matter.

Posted by Wile E Coyote | April 25, 2008 11:30 AM

Schools are fighting a losing battle. Many parts of society disparage education as something pursued by "elites" and elites are one step away from swarthy terrorists. In urban cultures book smarts and good grades are criticized as "acting white". Anti-intellectualism is a multifaceted movement unto itself. It truly is appalling what a great many 17-18-19 year old Americans don't know and don't care to know. History, grammar, the basic concepts behind electricity, hydraulics, pneumatics, carpentry, astronomy, meteorology. The workings of government, how a law is proposed and passed, the role of each our three branches. Have any of them ever diagrammed a sentence? Likely never. I had a 16 year old kid over to help around the house recently. His father was an acquaintance and the young man needed to earn a few dollars. I quickly discovered he'd never dug a hole with a shovel, operated a chain saw, dismantled a fence, swung an axe, used a paint sprayer, mowed a lawn, cleaned a gutter, pruned a tree or bush, used a sledgehammer or hammered a nail. WTF?! I'm sure he's somewhat typical of millions of kids raised in front of the TV or computer, either watching them or using them for video games. It won't be (as conservatives so tremble and fear)
immigrants or commies or gays or elites that cause us to crash and burn. We are our own worst enemy.

Posted by steve duncan | April 25, 2008 11:34 AM

Wile E Coyote,

I think most people who have studied these issues would agree that many other factors besides, say, spending per pupil on instructional costs affect the outcome for those pupils as measured by things like test scores. But that doesn't mean increased spending per pupil on instructional costs would have no beneficial effects even in cases whether those other factors remain unaddressed.

Generally, the proper question from a public policy perspective usually is not whether a given proposal would create an ideal outcome, because the relevant authority rarely has the ability to control all the factors necessary to produce an ideal outcome. Rather, the relevant question is usually whether the outcome with the proposal would be better than the outcome without the proposal, and whether those benefits can justify the costs of the proposal.

Posted by DTM | April 25, 2008 11:43 AM

Rather, the relevant question is usually whether the outcome with the proposal would be better than the outcome without the proposal, and whether those benefits can justify the costs of the proposal.

Posted by DTM | April 25, 2008 11:43 AM
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Obviously not the decision making model used in contemplating the war in Iraq.

Posted by steve duncan | April 25, 2008 11:49 AM

Here's Will on Moynihan from 2003, on the event of Moynihan's death:

The Senate's Sisyphus, Moynihan was forever pushing uphill a boulder of inconvenient data. A social scientist trained to distinguish correlation from causation, and a wit, Moynihan puckishly said that a crucial determinant of the quality of American schools is proximity to the Canadian border. The barb in his jest was this: High cognitive outputs correlate not with high per-pupil expenditures but with a high percentage of two-parent families. For that, there was the rough geographical correlation that caused Moynihan to suggest that states trying to improve their students' test scores should move closer to Canada.
For calling attention, four decades ago, to the crisis of the African-American family--26 percent of children were being born out of wedlock--he was denounced as a racist by lesser liberals. Today the percentage among all Americans is 33, among African-Americans 69, and family disintegration, meaning absent fathers, is recognized as the most powerful predictor of most social pathologies.

Posted by Al | April 25, 2008 12:10 PM

All of this misses the most important factor destroying American life today as heard on a variety of conservative radio shows from Michael Savage to Neal Boortz to many others: Tha Teachers' Unions! Apparently, every problem in life, economics and culture has been caused by them, and they make the Illuminati / Freemason / Knights Templar conspiracists quake by comparison.

Posted by El Cid | April 25, 2008 12:13 PM

steve duncan,

I have been both amazed and angered by the fact that what passed, and still passes, in many circles for cost-benefit analysis of the Iraq War basically just involves stating any conceivable benefits and ignoring the likely costs.

Posted by DTM | April 25, 2008 12:35 PM

FYI: Upstate New York is about one-third as cheap to live in as NYC, the DC area, Boston area, etc. It's more like the mid-west in terms of cost/quality of life issues.

Will miseads readers into believing Moynihan would agree with his rabid anti-public schools rants. The Schools Suck crowd has no real data to defend their decrepit positions.

Posted by maxgowan | April 25, 2008 12:45 PM

While E. Coyote - The leading determinant of school quality as measured by standardized test score is the proximity to... white people. And there are higher percentages of white people in the states that border Canada (and those close by, like Iowa) than those that do not.

True. Though it doesn't have to be "more whites the better" when looking at scores. Asians are as good or better than whites in academic achievement, it's just their impact is diluted in test performance scores in regions with high black low achievers in the same system. If not for Asians and a certain percentage of motivated Hispanics, the scores in certain black-heavy MD, VA, NYC, and California districts would be as horrendous as in other black-run cities and school districts.

(Bringing up the flip side of Moynihan's argument. The closer Canadian schools were to the USA, the better, until the recent wave of low-IQ black and Muslim immigrants to Toronto and Montreal brought academic achievement down. However, in Asian-heavy Vancouver, students beat most mainly white districts.)

While E. Coyote - The dollars = quality outcomes is simply false. Drill down futher than the state level to see if this is true. Throwing dollars at urban schools in Detroit, Chicago, Milwaukee, Minneapolis, etc. does not correspond to higher achievement. The outcomes based on race are profound, even within the same school district:
http://www.wpri.org/Reports/Volume%2020/Vol20No6/Vo20No6p1.html

True, gold plate a lump of shit, wrap it in hundred dollar bills, or a pricey teachers union that is to "nurture" the shit into growing into perfume - you still have a lump of shit.

If one looks at dollars spent per hour taught, America with a shorter school year spends more per capita than any advanced nation in Europe or Asia but Switzerland. The white and Asian performance is actually quite good outside the South and in Math and Sciences that are "dumbed down" so their dumber, less-motivated counterparts don't drop out. The big problem is the black racial disparity, with the Latins also showing a disturbing trend in 2nd, 3rd Gen to adapt black values and have a lower commitment and value on education.

While E. Coyote -The higher dollars per student spent in the northern states is a function of stronger teachers unions and a greater willingness/ability of the citizens to endure higher local taxation.
The quality of the product depends on many other factors, money being somewhere down the list from parental involvement and parental giving-a-damn. Without that, the money just does not matter.

True. The teachers unions dominate NE Democratic politics, and the Democrats dominate state legislatures and rubber-stamp teachers demands that money be poured down the public school ratholes, and that most of that go into "noble teachers out to transform lumps of shit into great students" salaries, bennies, and peks like "lower classroom size to inspire the uninspired".

Add to the parents of inner city kids not having a damn about education the "little darlings" themselves who are cursed with low-IQs, poor habits, violence, sloth, and who "direct crabs in a bucket" peer pressure against good students who work hard.
And of course, once ernest, well-motivated liberals who after putting 100% into their inner-city charges and attempting to uplift them, realize the problem is hardly with teachers, more money needed - it is with who walks into the school as a student and their crappy parents and communities. So, embracing the futility of changing lumps of shit happy and proud to be pieces of shit just like their Momma and whoever BioDaddy was - those idealistic liberal stalwarts conclude they must embrace futility, mediocrity, and do bare minimum - or if they still have a spark of ambition, get out and find a district inhabited by motivated students and parents.

Posted by Chris Ford | April 25, 2008 12:48 PM

Re RKU

Although the quality of education in DC schools is poor, the quality in some of the DC suburbs is among the best in the nation. In particular, Montgomery Co., Md., Fairfax Co., and Falls Church City in Virginia.

Posted by SLC | April 25, 2008 1:33 PM

Moynihan didn't say anything about per-pupil spending. Will just made that part up. He did say that the biggest factor in test scores was proximity to the Canadian border, which was pretty funny. At the time he made the joke, the first state by state comparisons came out, and North Dakota, Minnesota, and New Hampshire were at the top. And they're all near the 49th parallel. Get it? Another joke popular at the time was, what should California (one of the lowest-performing states) do to raise test scores? A: annex Oregon and Washington.

We educators are such cards.

Posted by Bob | April 25, 2008 1:56 PM

SLC,

"Although the quality of education in DC schools is poor, the quality in some of the DC suburbs is among the best in the nation. In particular, Montgomery Co., Md., Fairfax Co., and Falls Church City in Virginia."

Why is that?

A. That's where the involved parents who give a damn live.

B. These schools spend more per capita, and this additional spending explains the difference in outcomes between their schools and the ones in D.C.

C. Other

Posted by Wile E. Coyote | April 25, 2008 2:49 PM

Guys,

You do realize that Senator Moynihan (D-NY) was making a joke, right? Before he became a politician, he was a social scientist, so he'd studied the question endlessly. He just couldn't come out in public and say the real reason test scores are higher in states near the Canadian border, even though everybody privately knows the real reason:

Hockey!

Playing hockey makes you smart.

Posted by Steve Sailer | April 25, 2008 3:27 PM

So, I propose the term "The Hockey Gap" as the answer to all questions that might get anyone James Watsoned.

Posted by Steve Sailer | April 25, 2008 3:44 PM

I'd guess this very same map could be used to reflect rates of cigarette and alcohol consumption and percentages of obesity in adults, inverting it so that light colors reflect high rates of each with darker colors showing lower numbers. Yes, there'd be notable variances but I think it would be mostly accurate.

You're wrong as far as alcohol is concerned. It surprises a lot of people, but per capita alcohol consumption is actually pretty low in the South. Where is per capita alcohol consumption the highest? You see that very dark state between Minnesota and Michigan? Yep, 'Sconies love to drink. And overall, people in the Upper Midwest consume far more alcohol per capita than people in any other region of the country.

Posted by asdf | April 25, 2008 9:31 PM

Anyone who uses the word "puckishly" without irony should be strangled with his own bow tie.

Posted by SqueakyRat | April 26, 2008 3:34 PM