GothamSchools — daily independent reporting on NYC public schools

Ken,

1.1 million students in NYC public schools, right?

50,000 applications to charter schools, right?

So, less than 5% of student (or families or something like that) are dissatisfied enough with the traditional public schools to even apply to charter schools?

Well, I would not argue that. But don’t those people who point to the popularity of charter schools as an major argument for the value (or something) have to account for the fact that more than 95% of students do NOT apply to charter schools?

I can’t see any objective reason to agree to that.

The fact is that we have schools in this city that have families applying to them from all boroughs. The school choice model is city allows families to apply to schools outside of their neighborhood. No doubt that some of the 50,000 who have applied to these charter schools live *outside* the immediate neighborhoods of these schools.

So, if the numerator includes students from outside the neighborhood, why shouldn’t the denominator?

Furthermore, placement of charter schools is hardly an exogenous issue. We would see more charter schools in other neighborhoods if parents in other neighborhoods were that dissatisfied with their schools. Truly understanding the degree of demand for charter schools and the potential of the NYC DOE requires looking at the whole system. If you were evaluating the performance of an investment fund, you would not just look at the good years or just the bad years. You would just look at the rising stocks or just the falling stocks. You’d look at the whole picture.

It would be fascinating to look at the % of students in each neighborhood who go to private schools, go to public schools and go to charter schools (with, of course, a listing for the % who applied to charter schools but did not attend, for whatever reason). But it would be blatantly dishonest to claim that all applicants come from the immediate surrounding neighborhoods, especially when we already know that families in this city are willing to send their children across the entire city to get them to the right school.

Only if living “nearby” was a requirement for applying would it make sense to limit the denominator like that. In fact, I’ll bet that if you were able to exam the zip codes of the applicants, you would find plenty from other neighborhoods. And if you didn’t, I would have to suggest that the best explanation is that potential applicants (and their families) do not think that these schools are worth the long commute — unlike the city’s best non-charter schools, public or private.

Ken,

Doesn’t New York City show that parents will send their kids virtually any distance to get them to a good school? How much ought you weight distance? What data would you use to figure out the weighting function?

I have the feeling that you might elect to use current applicant data at these charter schools, but I think that its fairly obvious that you should NOT. We already have schools with unquestionably excellent reputations in this city, and they are ones that demonstrate what families will do for excellent schools. These great private schools and great public schools might have different applicant patterns that these charter schools do, but that only serves to illustrate a smaller gap in perceived quality between charter schools and traditional public schools generally than between (for example) the specialized exams schools and traditional public schools.

(Obviously, families are usually less willing to send their primary or middle school aged children across the city by public transportation, so different weighting formulas would apply at those levels. But you could still look at applicant data at the city’s most in-demand elementary schools. Heck, perhaps you could look at the city’s most in-demand preschools to get unbiased look at the distances parents are willing to travel to take their small children to and from school?)

I think that you don’t follow my single paragraph because it goes against your instinct.

You tend to think that local neighborhood should be weighted much more heavily. I have trouble justifying that.

1) In most circumstances, I would agree to at least some degree, as transportation really is a big consideration in most areas. But our public transit system in this city largely take care of that.

2) Private schools have long shown that people are willing to take their kids quite far from their neighborhoods — even without public transit — to get them to the “right” school. Specialized exam schools in this city show that parents and students are willing to put up with transit times of well over an hour, to get kids to the “right” school.

3) I have no doubt that the value that parents put on education has an impact on how far from their own neighborhoods they are willing to send their children to school. The charter schools in this city are NOT concentrated in neighborhoods with high educational attainment. It would follow, then, that families in other –even far distant — neighborhoods would be willing to travel all the way to these schools, if they are the “right” schools.

There simply are SO-OOOO many argument as to why we should include a very broad segment of the city in the denominator — including families who send their kids to private schools, of course — and rather few that that actually hold up even to weight distance very strongly as a discount.

The fact is that charter schools serve a *small* minority of the city’s families, and there is a relatively *small* demand for them. One has to do dishonest and/or biased things with the math to show that they are anywhere near as popular as many charter school backers paint them. There are many other much more popular ideas — both in and out of education — that we never try, but for some reason we keep doing this charter school thing as though governance changes will impact instruction in the classroom.

Why?

(And you’d already acknowledged the problem of the numerator double-, triple- (or more) counting many families. I didn’t feel like I needed to belabor it. I have plenty of my own points to belabor.)

MM,

Sure, they would prefer a quality local school. We agree on that.

However, the question we are trying to examine is how they judge the quality of their local schools vs. the quality of other public schools vs. the quality of private schools vs. the quality of charter schools.

One of the arguments against charter proponents is that charter schools’ popularity is an indicator of their quality, the superiority and why they should be expanded. Some might use the raw number of applications to charter schools to try to makes these points, without regard to the size of the population from which they are drawn (or, as Ken has pointed out a couple of times, the number of charter school applications for each applying student). There was a time when people pointed New Kids on the Block sold albums faster than the Beatles did, without accounting for the larger population, of course. We need to consider context.

I think that this relatively small number of charter school applications demonstrates that most families do NOT think that charter schools are notably superior to their other options — including their local public schools.

As for the student Metrocards….well, Ken and I are looking back on this stuff, so their possible axe is irrelevant. Looking forward, it would be hard to say. It would be interesting to compare the applicants’ zip codes to schools last year (before even rumored axing) to this year (potential axing) to next year (either axed or not). Personally, I am hopeful that the politicians and policy-makers will figure THIS one out. Either way, it would be an interesting study.

MM,

Well, that would be an interesting comparison, certainly. Let me grant that it could be really interesting. However, you’d have to adjust/account for the various kinds of school choice available to parents. To really understand this, you’d want to compare to other forms of choice schools and schools in other neighborhoods. It gets really complicated, because NYC does not have the kind of simple catchment based assignment of schools that so many of us grew up with. There are so many different things going on (e.g. catchment, choice, private, charter) that doing the comparison right would take a LOT of planning.

Of course, high quality quantitative research times a LOT of work on the front end. The more interesting the system, and the more concerned you are with getting meaningful and useful results — as opposed to weak evidence to support ideological positions — the more work it takes.

As for Avatar…there’s a lot we don’t adjust for. There’s population, inflation, and countless others. And then there’s the various accounting tricks (See the new book, “The Hollywood Economist). There are a lot of reasons not to trust claims about Avatar’s historical place, based on it’s take at the box office. I think we agree on that.

OK. RttT is so much less than the war effort in Afghanistan. But how does total federal spending on k12 education compare? What about total governmental spending?

Why is your comparison the appropriate one to make? It seems like you are comparing an apple to differential calculus. (Well, it’s probably not THAT bad of a comparison, but it’s pretty bad.) In what context do you think that this is the right comparison tom make?