Schools and Officers’ Wives

I’m reading this fascinating interview in the Financial Times (possibly paywalled) with my former super^n-boss Lloyd Blankfein. It’s full of interesting nuggets, as well as fodder for people who want to criticise him.

I must admit right up front – I’m a big fan of Lloyd’s. This has nothing to do with the fact that I briefly worked for Goldman when he was CEO (I had even asked him a “planted question” when he had given a talk to the office sometime during my tenure there). In general, I think he says things as they are, and his twitter account is rather entertaining as well.

Anyway, the first statement in the interview that caught my attention was this statement about why the quality of schooling has gone down over the years. “He explains that the schools were only good because the women who staffed them were blocked from jobs in business and industry.” This is complementary to a view that I’ve strongly endorsed for a while.

Let me explain this using examples from India. Long long ago, maybe until the 1940s and 1950s, most school teachers in India were men. Way too few women had the kind of education that would qualify them to teach in schools. Moreover, back then, teaching paid sufficiently to run a (at least lower middle class) family on a single income.

In the 1950s and 1960s, women in India started going to college, and started entering the workforce. Mind that it was still a massively patriarchal society here back then, and women were expected to do their “household duties” in addition to bringing home a secondary income. And this meant that many of them were in the market for jobs that offered good work-life balance.

Teaching in schools offered that sweet spot – it required credentials, and the woman’s degrees would help in that. The hours weren’t too long. There would be ample vacations through the year. Schools were found everywhere, so the job was location-independent to a large extent. This last bit was important since the women’s husbands would frequently be employed in government jobs that were transferable, and the women’s “secondary careers” meant that they would be forced to move along.

And so we saw the rise of a class of teachers that I’ve come to call (not very politically correctly) as “officers’ wives”. These were well educated women, married to well educated men who held good jobs. They were passionate about their jobs, and went about it with a sense of purpose that went well beyond making money. This meant that the standard of teaching overall was raised.

And most importantly, this increased standard of teaching came without a corresponding increase in cost. The marginal utility to the family of this secondary source of income wasn’t particularly high, so this class of teachers didn’t demand very highly in terms of wages. In any case, they were doing their job out of passion rather than for the money, and would be willing to accept below-market wages to go about their jobs.

Then, two things happened. Firstly, the presence of employees who weren’t in it solely for the money pushed down average wages, and teachers for whom teaching was the sole source of family income started getting crowded out of the market. Secondly, with liberalisation in the 1990s, the nature of the job market itself suddenly changed.

One reason why the “officers’ wives” took to teaching was that it was hard to find other employment that was commensurate to their education that gave them the flexibility they desired (if you’re a secondary income earner you need that flexibility). With the market opening up, there was suddenly a number of options available to these people that matched their skill and flexibility needs. For example, my 11th standard physics teacher quit the school midway through the year to take up a job as a software tester at Wipro (this was in 1998-99).

So, rather suddenly, the opportunity cost of teaching shot up since the teachers suddenly many more options. It wasn’t possible for schools to jack up fees at the same time to be able to continue to afford the same teachers. And so, supply of quality teachers dropped. And consequently, the average quality of teachers (holding the schools constant) dropped as well.

Putting it in another way, schools nowadays need to compete with a much larger and much more diverse set of employers for their teachers. Many of them, for the sort of fees they charge, are simply unable to do so. The “passionate bunch” has found other avenues to exhibit their passion.

And the problem continues. And from what Lloyd says, it isn’t only India that is seeing this drop in quality of teaching – the US sees that as well. It was a sort of repressed larger market that had artificially pushed up the quality of school teachers, and the drop in repression has meant that the quality of teaching has dropped.

I will leave you with the concept of Baumol’s Cost Disease.

Baumol Disease Index

In his excellent take on why Rohit Sharma’s 264 is bad for cricket, Niranjan Rajadhyaksha writes about the Baumol’s Cost Disease. This phenomenon, which was first described by William Baumol and William Bowen in the 1960s, describes the increase in cost of labour in industries that have seen little productivity. This has to do with an increase in productivity in other sectors which pushes up the clearing price of labour, which increases the costs of industries that have seen no improvements in productivity.

Based on this, we can construct an index on how industrialised an economy is, which I’m going to christen “Baumol Disease Index”. The basic idea is to pick a sector that is likely to be unaffected by productivity changes over the long term, and look at the median salary of workers in that sector in different countries and across different points in time. This can help us compare the relative levels of industrialisation and productivity in different countries, and in the same country over time.

In order to construct this index, we will take into account one sector which has a lot of “human input” and is unlikely to see much improvement in productivity thanks to mechanisation. My first choice for this was for employees of a company like McDonalds (taking off on The Economist’s Big Mac Index) but then that sector is not that insulated from greater productivity.

We could use the original example that Baumol and Bowen used, which is performing arts, but then performing arts is a winner takes all market – Iron Maiden will be able to command much higher ticket prices compared to the local orchestra thanks to their history and brand and perception of quality. So performing arts is not a great example, either.

Another good choice would be government bureaucrats, since their work is unlikely to be much affected by productivity. But then we’ve had some computerisation and that must have increased some productivity, and ability to be productive and willingness to be productive don’t always go hand in hand!

What about drivers? Despite the efforts towards development of driverless cars, these are unlikely to really take off in the next couple of decades or so, and so we can assume that productivity will remain broadly constant. The other advantage of drivers is that while salaries are tricky to measure (and we need to depend on surveys for those, with mostly unreliable results), taxi fares in different cities are public information, and it is not hard to separate such fares out into cost of fuel, cost of car and cost of driver’s time. This way, measurement of an average taxi driver’s income in different cities and countries, and at different points in time, should not be really difficult.

So, I hereby propose the Baumol Disease Index. It is the per month pre-tax expected income for a driver in a particular city after taking into account costs of fuel and car. This number is going to be imputed from taxi prices. And is going to be a measure of general levels of productivity and industrialisation in an economy. Sounds good?

And while we are on the topic of indices, you should read this excellent leader in last week’s The Economist on the profusion of indices. And since we have a profusion anyway, adding this one additional index shouldn’t hurt! And this one (Baumol Disease Index) measures something that is not measured by too many other indices, and is simple to calculate!

Howzzat?