Tag Archives: education

Edit: I just realized there’s a similar argument from earlier today arguing for public schools as a “moral adjustment”. I’ll have to think more about this, but it seems pretty unreasonable expecting parents to send their kids to the vastly worse public school without expecting that all other parents of their class will do the same. Law is the only way.

A lot of people on Twitter today are arguing about the merits of government policy that advances public education by taxing private schools. That’s a bad idea. We shouldn’t tax private secondary institutions – we should get rid of them.

I’ve discussed this idea before, but it’s worth considering in terms of externalities on the margin. Normal discussions of elite educational privilege juxtapose the yearly expenditure per pupil at a public school in Camden against Andover, or something.

This is not the best metric. While many failures at the local level may be traced to a lack of funding, institutional arrangements are far more important. By this I mean, kids of smart, working professionals ensure a good education for their students not only with a higher tax outlay, but by constantly demanding teachers to live up to their potential.

In fact, you’ll find teachers in affluent suburbia complain about the over involvement of affluent parents in their teaching process. I can remember both my parents dealing with pretty big problems at my pretty good public middle school and very good private (international) high school. (My parents are actually huge believers in public education. Except even they’re not nuts enough to send me to such in India).

That teachers are constantly held to a higher standard institutionally forces the administration (which was, granted, decidedly shitty at my high school) to bring in better teachers and provide a better curriculum.

Rich parents will stop at nothing to make their kids’ education world class. Now let’s say the government passed a law forcing me – and ten similar friends – to go to the crappy high school across the river. They also passed a law preventing us from using private educational services to bypass the public system.

Even if the school in question received no more public funding, our parents would do their best to make sure the education was of a higher quality. Not immediately, but eventually: through more conversation with the administration and perhaps even private charity. The presence of rich kids in a poor class room pays huge dividends to the disadvantaged kids in the classroom.

At a microscopic level, this experiment is almost bound to fail by bureaucratic paralysis. But, more broadly, if a new law required that all kids may be randomly assigned to any public school in the area, and private schools were banned, affluent parents would ensure all schools that their kid might end up at were the best possible. They would vote their own property tax rate up, and finance huge fundraisers. They would have no other choice, lest their dear little son end up at a bad school.

There is huge public gain for every marginal transfer to this effect. Banning private schools would be necessary to prevent the rich from opting out of the system altogether. It’s also unlikely this would hurt rich kids too much, since affluent parents, we might say, are completely price-inelastic when it comes to the quality of their kids education. It just means less money for a ski vacation. Boo hoo. (Actually, if you subscribe to the public good externality hypothesis that I do, even this would be pretty small. But rich white and asian parents also experience disutility when their kids interact with lots of blacks wearing hoodies, so who knows about the welfare loss there…)

This would eliminate any need for broader governmental subsidies, and might even make obsolete the Department of Education. This is illiberty, but so is taxation, and so is the minimum wage. I’m not appealing to the libertarians who want an end to all government but the liberals who think some forms of illiberty are better than others. This is no different.

The best way to help society is to align the private incentives of the rich with the public goals of the poor.

This one’s heavy on the “radical” and light on the “centrist”. 

I knew I had a bit of Marxist in me somewhere. There are remarkably few instances where elimination of private services can have any benefit. Some recent (broadly unrelated) findings from Raj Chetty, Nathaniel Hendren, Patrick Kline, and Emmanuel Saez, to me, suggests otherwise. But I’d argue my following proposal for education reform is also sufficiently centrist as – given certain important caveats – it will heavily diminish the Department of Education’s driving mandate.

Consider a country where the federal government didn’t fund any educational institutions. It levied no redistributive taxes to help the poor catch up with their richer brethren. We could well be living in a night watchmen state. It did, however, institute one law: private education is banned at the primary and secondary level, and federated states must randomly allocate children into schools across a reasonable geographic distribution. Another way to think about it is all privately established institutions (aside from a select, grandfathered few) must offer entry by lottery (this would also deal with a few adverse selection effects of which I speak later).

Socioeconomic cohesion is a natural argument in favor of this proposal. But the potential upsides run a little deeper than that. Take a very divided city like Chicago, with a rotting inner-city populated by blacks and gentrified suburbs with private and what-might-as-well-may-be private schools catering to the affluent. In Chicago, like the United States around it, the merely middle class and affluent do remarkably well. In fact, for families earning over $50,000 – a healthy, but definitely not incredibly high, sum – Americans are amongst the most educated internationally, contrary to stereotypes otherwise.

Lagging education in America derives largely from a large number of poor immigrants and a historically impoverished underclass. (Apologists of our system rightly note that adjusting for ethnicity, Americans are the amongst the best-educated youth in the world).

Here’s the immodest proposal. We know that middle class and affluent schools are doing something right. We don’t know what, and we certainly haven’t been able to export that know-how into success at the inner-city and rural areas. Throwing money, by itself, hasn’t helped either. There’s good reason to believe a lot of it has to do with community engagement which thrives when parents can afford to take a lot of time off to hold the school accountable, and have the confidence to fight against a bad administration.

But let’s say all kids in Chicago were randomly allocated to a school within suitable distance – long enough that rich kids might have to go to poor areas, and vice-versa. The large number of affluent plus families in the area would be compelled to ensure all schools are up to snuff, lest their poor kid end up in a bad program.

This might mean that the rich would voluntarily raise property taxes to finance local education – because private schools are illegal it’s, to their chagrin, the only way to increase financing. It might also mean that schools cut back on wasteful or unnecessary programs. Whatever the case there’s little reason to believe resource scarcity will be a serious problem in a country as rich as the United States.

Because public education is a very localized process in the United States, we might say there’s a “wisdom of the market” that lets good schools succeed. That is to say we don’t know exactly what they do, but if we align those with resources to the right incentives it’s not hard to replicate. We can bank on the fact that America’s extremely successful upper middle will never let their children go to a shoddy school. Today, because richer districts pay well and offer better working conditions, the affluent schools also crowd out the labor market for teachers leaving less motivated and competent teachers for the poor. This proposal would paralyze that competition.

The big caveat is that if states randomly allocate kids across a narrow geographic area, adverse selection processes will compel the rich to select themselves out into their own little cove to a greater extent than today. This isn’t a light issue, but there’s no reason to believe that geographic residence “markets” are perfectly competitive and that people want to gentrify beyond a certain extent (the allure of living in exclusive neighborhoods is probably as conspicuous as it is utilitarian).

We haven’t seen high taxes in richer, bluer states like New York – Rick Perry’s Texas campaign be damned – drive out the best and brightest. In general, it’s a lot easier to move within a locale than between locales, and the evidence that we might have broad, cross-regional assortative effects is extremely weak. (Otherwise California and New York would be poorhouses, by now).

I think this would elevate educational attainment and lower expenditure as a percent of GDP, at the same time. Aside from paying teachers more – which richer districts do quite well – a lot of the money we’ve thrown at education has been for naught (many, much poorer, countries do better for the poor than we do). By getting government out of any and all educational subsidy, the random allocation mandate would force an efficient utilization of scarce resources – something that economists would love.

Naturally, this is a truly radical proposal that murders any resemblance to our deeply unequal, yet government-dominated, educational system of today. I’m not even sure I support it but, given a few caveats, it could offer important insight into making education work. It’s a deep abridgment of personal agency and liberty. But it’s also liberating. Whether we like it or not we live in a society with a strong government omnipresent in most activities. It’s unclear whether this is any the more illiberal than taxation or regulation, but it has potentially grander benefits.

Evan Soltas has a very interesting blog post about intergenerational inequality. The thrust of his argument is that inequality in educational attainment exists, even adjusting for differences in income. At some level, this is trivial – one would expect educated parents, wealthy or not, to spend more on academic pursuits for their children. The conclusion Evan draws, however, is quite fantastic that is, the effect of educating a kid today is geometrically valuable:

Here’s why that math is important. The high school dropout rate among people whose fathers were dropouts is 22.2 percent. The dropout rate with high-school-grad fathers is 2.9 percent. Let’s assume that the social value of a high school degree is $30,000 per graduate; that’s roughly the difference in average income between non-grads and grads. Public policy that supposes they are helping one person assesses the value of that degree at $30,000, obviously. Public policy that supposes they are helping an infinite succession of people assesses the value of that degree at $819,000.

He arrives at a whopping $819,000 as the difference in the net present value of fathers with and without high-school degrees in terms of the future earning potential of their kids. I would argue that the value of a high-school degree is somewhat less than $30,000 because a good amount of the delta is derived from the further division between college graduates and non-graduates. This wouldn’t change the situation if the deviation was randomly distributed (and college actually added value, a contestable claim), but there’s good reason to believe it’s not. (For example, there are definitely demographics in which it’s very likely a child graduates high school, but also very unlikely that he or she graduates a 4-year college).

That even basic jobs require a college degree nowadays would further diminish the value of a high school degree by itself. The $819,000 figure includes the increased chance of attending college, which is far more dubious in its actual value added. A more conservative, but not necessarily better, approach would be to estimate the value added by high school itself, assuming that college adds no value, or approximately $11,000, according to Pearson.

The geometric multiplier is about 26.15, yielding a present value of $298,615. This isn’t nearly as astonishing as Evan’s $819,000, and probably isn’t as close to the truth. However, it does compound an already compelling case for investing in high school education; that it’s worth it even if no one ends up at college. At a public school, the average cost per pupil is $10,652 or about $40,000 for a degree, or a 645% return on investment. This is a low-hanging fruit.

I think there’s one big flaw in the way Evan looks at things; it’s also important to note that the value of a high school education is hugely understated by the calculated figure, because it is the difference in annual income. The effect over a lifetime is many multiples of this. For this reason, the return on a high school degree, a net of 645%, doesn’t begin to approximate the real value of an education.

A final caveat, I have a pretty strong feeling that education by itself doesn’t explain the intergenerational inequality in education. As it happens, I came across this passage in a recent copy of the Economist:

The well-to-do [have] largely held to old-fashioned ideas about marriage. Among professionals, births within marriage are four times as common as births where the father is registered as absent from the household.

This is for Britain, but the gap is only worse in America. The figure below charts educational attainment of a child by family situation at age sixteen.


One can see an almost twofold increase in dropout rate between dual and single guardian households (the exception being families where neither of the guardian is directly related to the child). The geometric multiplier between a nuclear family and anything but is about 20. While this is less than that of a high school degree, the policy cost is also substantially less than $40,000.

Today, in many ways, we disincentivize marriage – as Nicholas Kristoff grimly notes – with poorly designed welfare programs (such as handouts to single mothers). As the Economist notes, I don’t believe the chain of causality is as strong between an educated parent inculcating similar values in his children as much as strong tendencies towards the nuclear family among the educated, which themselves engender stronger attitudes about education.

Regardless, any of these points are a damning indictment of the “American dream”. Policymakers need to acknowledge the importance of a high school education. We’ve spent billions in grants and subsidized student loans for college that implicitly benefit the lucky, that is those who already have a high-school education.

Further, it is far from established that college today adds economic value to society. The huge underemployment of recent graduates (which itself causes the unemployment of those without a college degree) should make each of us think twice while promoting the virtues of college.

Indeed, the advent of online courses further convinces me that college is but a licensing factory. For example, a University of Washington course in Financial Econometrics is provided on the Coursera platform, with the additional option of doing slightly more work and receiving slightly more attention, at a price tag of $4,000. This is absurd. The only value derived from the for-credit option is Washington’s official stamp of approval, a license.

By blindly subsidizing college education we’re implicitly encouraging rent-seeking behaviors that are damaging to society as a whole, especially those at the bottom. Evan’s point is clear: we need to stop acting like a high school degree is its present value, so long as inequality persists. Indeed, Evan hugely underestimates the value of his own claim by ignoring the lifetime effect of increased cash flow. For those of you interested in that figure, even discounting at 15%, it’s over $10 million.