My profound confusion about urban schooling follows from a talk, a website and a conversation with my colleague Partha.

First, the talk. An NGO based in Delhi operates a school for children in slums. For the longest time, they have tried to “register” the school so that the children can participate in central examinations and receive an appropriate school-leaving certificate. They were unsuccessful. In the end, they had to go through the “open schooling” system, certificates from which are discounted on the labor market. The reason for their failure was the elaborate system of regulations and rules under the Delhi Education Act of 1973.

The act requires lots of paperwork, schools that have playground facilities, teachers who are paid “an equivalent amount to teachers in schools run by the administration”, laboratory facilities, a library that is well stocked—you get the picture (one gem is the bit on corporal punishment, where head-teachers are allowed to cane erring students, but the caning has to be restricted to “10 strokes on the palm of the hand”).

Next, website #1. Go to the wikipedia entry on Delhi Public School, Mathura Road. As the entry assures us, the school is built on a “sprawling” 15 acres of land given to it by the government in 1949 and coincidentally, is right next to one of the most posh localities in Delhi. Here is a simple calculation: multiply 15 * the price per acre. That’s what you get if you sell the land today. Divide by 10% (the interest rate on a one year CD in India) to get the annualized value. Divide by 4000 (the number of students) to get the annual per-student subsidy from the school. The result is a staggering Rs.700,00 per year, or close to Rs.60,000 ($1500) per month. That’s in a city where the income per capita is less than $1000 a year.

My thinking: There is no way that a private school is going to come up in this area without government subsidies at that price. In fact, the subsidy does not make sense even if you take into account estimates of the returns to education.

My first source of confusion. Comparing subsidies in this way is wrong, since you have to account for land appreciation as well. One way in which economists deal with asset appreciation is to follow Hotelling’s rule, which says that you should compare the interest rate you get from selling it today and keeping the money in the bank to holding on to the asset and selling it tomorrow at a higher price. Getting back to the school, there are now two choices:
a. Tell students that the school is being shut down and they will get Rs.60,000 a month, but have to get educated somewhere else. Most parents would probably take the offer—a top-notch private school outside the city costs around Rs.20,000 a month.
b. Taking land appreciation into account, you could also make another offer: “wait a year, and I will give you Rs.70,000 a month (yes, land prices are going up faaassttt) after closing down the school.”

I don’t have enough data to figure out which should happen, but neither is particularly satisfactory. Under option (a) all schools eventually leave the city; under option (b) there is a (probably) perpetual subsidy from present to future generations.

Finally, the conversation. As Partha points out, schools generate externalities, so the relevant question is “what would the land price be if there were (taking it to the extreme) there were no schools in the city?” In the U.S., school quality is incorporated into housing prices. Although in Delhi it’s only recently that legislation has been passed that requires a “close-to-school” residence requirement, it’s probably true that land prices would drop as schools migrated out. This was turning out to be an impossible problem, since empirical estimates for “land prices in the absence of schools” are non-existent and likely to remain so.

Fortunately, Partha also suggests a solution. Instead of looking at buying and selling land, why not look at land rents, which price in the appreciation of land over time? This makes life easier. Even in the posh localities of Delhi, land rents are around Rs.400 per square foot. Take 200 square feet for a classroom with 40 students and you get a much more reasonable price of Rs.2000 ($50 per month). Although not affordable to everyone, upper-middle income groups and the rich (who would send their children to schools in posh localities) could easily pay this every month.
 
The problem though is that a classroom is not enough. According to our government, a school cannot run unless it has a playground, laboratory facilities, a library, a staff room, a place for students to spit (!) and a host of other requirements. Neither is this a problem of only posh localities with high land prices—our speaker from the NGO assured us that if they tried to fulfill all the infrastructure requirements of the Delhi Education Act, it would cost them Rs.18,000 a year per child—that’s half the average per-capita income of individuals in the city, and about 3 times the per-capita income of people living in slums.

Although the luminaries who thought up the act probably wanted a holistic education for children in the city, they have ended up guaranteeing a tremendous supply crunch that can only get worse over time. So much so that an expert body had to recently legislate how schools could admit students (more on this later). Meanwhile, the only way to educate the poor is to stay under the radar, hope you don’t get noticed and live under the perpetual threat of your school being closed down.