Aiming for Mediocrity (archives)

Mark Kleiman had a good post about public education. I believe his main point, and one which I agree with him completely, is captured by this paragraph:

One of the many bad features of NCLB is that it focuses entirely on the bottom of the distribution. Our smartest students are being systematically cheated. That’s less true in rich neighborhoods than in poor ones, but it’s true everywhere.

In email follow up, Mark said that, “the logical implication of no child left behind is no child allowed to get ahead.” This is something true not only of NCLB, but of modern education policy in general. Since at least the late 1960’s the US public education system has poured funding into special and remedial education programs, to the detriment of gifted and talented students. Though I fully support special education, and have worked in that field, I realize that education funding is a zero sum game. Though we can argue and agitate for more education funding, in the end the pie is of a fixed size and every dollar going to the one student is not going to another.

This is a sensitive subject because of the prejudices people infer when you call for greater resources for gifted and talented students. This is especially so because up until more recent times, students with special needs were cast away by our public schools. And what I am arguing for is not a return to those days. But what I am calling for is an understanding that gifted and talented kids have special needs as well. Just as a slower learner needs assistance in keeping up with his/her peers, a more advanced learner needs materials and challenges to keep him/her occupied until the rest of the class catches up.

Until we recognize the needs of advanced learners and begin to provide resources for their educations, our entire system will be one that shoots for mediocrity. Just look at NCLB as an example. The focus is on bringing the slower learners up to some standard. Now, one can argue about what that standard ought to be, but it hovers somewhere below the mean. It would have to in order to ensure that a majority “succeeds.”

In education, and perhaps society in general, we are uncomfortable with inherent variations in intelligence. That intelligence is distributed on a bell curve (not to be confused that racist tome) makes us queasy. We want to believe that all children can get to the same place educationally, assuming we provide them with adequate resources. But in reality, this can never hold true. Intelligence is no different from athletic or artistic abilities.

These natural variations provide fertile ground to examine our values. Do we believe in equality of opportunity or equality of outcomes? If we believe in the former, then we should not be uncomfortable with some students achieving more. After all, we regard varying degrees of musical ability in children as unremarkable. Why is it that education drives us to regard unequal outcomes as inherently bad?

Two possible reasons come to mind, one rational, the other emotional. On the rational side, our country has a history of pervasive racism, more than a dose of sexism and institutionalization of special needs children. This fuels a perception that the variation in outcomes we see is the result of socio-cultural factors. Maybe we believe that if not for those factors, variation would be minimal. Or, we simply feel that intelligence is not normally distributed. Maybe we are unwilling to accept that reality.

No matter what the driving force behind our belief in equality of outcomes, we must resist that urge. There simply is no possible way to achieve that goal. Not only that, but many strategies aimed to reduce variation are actually attempts to enforce mediocrity. In our attempts to lift the boats of those less gifted we have abandoned all of those whose ships are not only floating, but cruising.

(originally posted March, 2007)

Advertisements

Tax Salience and Human Capital Investments

A new paper by Guvenen, Kurusku and Ozkan on income inequality and progressive taxation makes some interesting, though far from novel in the economic literature, claims about the effects of progressive taxation on human capital investment decisions. The theory is that progressive taxation distorts people’s incentives to invest in human capital, whether that be on the job training or returning to school for further education. The authors claim that this underinvestment in human capital is what lies beneath the slower wage inequality growth in the Continental European countries than in the United States. And, while their model is a sound one, I have vigorous objections to this particular conclusion.

Leaving aside a more general indictment of rationality in economics, I would like to make an argument about salience and then circle back to the issue of rationality, where I will claim that factors completely removed from taxation underlie much individual decision making with regards to human capital investments.

It’s no real surprise, or it ought not be, that tax salience will affect people’s response to a particular tax. The more visible the tax is, the more impact it will have on individual’s decisions to spend or invest. In this important paper, Chetty, Looney and Kroft provide experimental support to the notion that tax salience matters. Though this particular paper focuses on the effects of sales taxes and consumer purchasing decisions, it is reasonable to believe that its conclusions apply also to the impacts of other taxes.  (More evidence is provided in the field of public finance- there is a reason why governments are more willing to raise fees than to raise taxes, and that is salience.)

What does this mean to the effects of progressive taxation on human capital investment? Certainly income taxes are relatively visible; people are obviously aware of them. So it would seem to follow that they have a high degree of salience. However, I would argue that their salience is actually fairly low. Generally speaking, people are unaware of their marginal tax bracket, let alone fully understand how marginal taxation works. How many people, if asked, could say what their total income tax had been in the previous year? Very few, indeed.

Thus, income taxes, while very well known, have a small amount of salience. It is almost as if they exist in the ether. And, as has been shown by Chetty et al, low levels of salience will not distort decisions in the same way more salient ones will. It is therefore difficult to believe that a vague understanding of possibly higher taxes in the future accounts for much underinvestment in human capital.

Then what does drive individuals’ decision to invest in human capital? I would argue that there are a host of factors at play, and that they vary among individuals. For some the decision to invest is motivated by an inherent value placed on learning. Others value learning and education, but only as an instrumental value toward furthering their prestige, status or future income.

There are an almost infinite number of factors that influence the decision to invest in human capital. And, for some people (the super-rational), that may include weighing the costs of progressive taxation of future income streams that could be gained by such an investment. But for the vast majority of people, future rates of income taxation are just not salient enough to affect their decision one way or the other.

A Risk Worth Spreading

It’s no secret that the bank bailouts were largely a way of socializing risk while largely keeping the rewards in private hands. People on both the political left and the political right have raised legitimate concerns about these policies, with the elites, by and large, turning a deaf ear to the masses. But there is a risk that I believe everyone concedes needs to be spread over a much larger population. That is the funding for special education.

As the system now stands, local education agencies (LEA’s) are responsible for providing special education services to the students who reside within their district. Although a large portion of these services are not prohibitively costly, many are. This puts school districts in the difficult position of balancing requirements contained within SPED laws and their regular education services. LEA’s do receive state and federal dollars to offset some of the costs of SPED, but no one believes these funding streams are adequate.

A more just system would diffuse the costs over a much larger population. In the case of the US, that would mean allocating the costs of SPED to the federal government. This would ensure that LEA’s would not be forced to choose between providing appropriate education to its SPED students or its regular education students. In reality, there is very little choice. Schools often shortchange both parties in an attempt to split the baby.

How would such a system be structured? First, the only costs that ought to be borne by the federal government should be those above a certain percentage of the district’s per pupil cost. A system that merely paid out 100% of all costs above district average per pupil runs the risk of incentivizing LEA’s to over-provide SPED services, as they are not facing any of the increased costs. A reasonable starting point might be 150% of average per pupil, though this might be too low. It should be set at a level such that LEA’s are not facing exorbitant residential placement costs, but not so low that costs for more routine services are passed on to the federal government.

In addition, some sort of review process must be in place so that LEA’s are making reasonable decisions with respect to the services included in a student’s IEP. We should be careful that a student who might succeed in a mainstream classroom is not pushed off into a residential or other outside program, merely to reduce district enrollment and effort. This is probably the most difficult portion of the policy to get right. I’d imagine some sort of delegation by the US Department of Education to either state education agencies or some other contractor to provide evaluation services. But it still needs to be a process that respects the rights of the child and her family, such that an appropriate education is being provided.

Ideally, the payment system would be structured as a reimbursement as opposed to a grant. This would ensure that LEA’s are legitimately providing the services and placements. Given technological advancements, such reporting of claims for reimbursement would not place an excessive burden on LEA’s. It is also important that claims are paid within a reasonable time frame such that LEA’s and/or the municipalities in which they’re located are not disadvantaged, vis-a-vis cash flow.

None of these ideas are particularly new or even novel. We have known for decades that SPED funding is poorly designed. And that this structure needlessly pits SPED children and their families against LEA’s and regular education students and their families. There is simply no good reason for continuing such a flawed system.

Higher Education- Sorting and Signaling?

I want to riff off an interesting post by Tyler at Marginal Revolution about pedigree bias in economics departments. To give a brief summary, the top 30 economics departments hire almost exclusively from 6-10 PhD programs. The comments are interesting (read them) and there are a few themes I want to touch on here.

Underlying this entire discussion, for me, is this question- what are the goals of education? I mean this in its broadest sense, to include primary/secondary as well as post-secondary education. There are two primary purposes that instantly come to mind- preparation for the workforce; preparation for civic participation.

We can think of preparation for the workforce to include such things as discrete skills, such as those related to the use of technology to craft specific tools, such as principles of accounting. There are, quite obviously, a near infinite number of skills/tools that are direct preparation for a career. It makes little sense to attempt to enumerate those here.

Next, what do we mean by preparation for civic participation? I would argue that what is meant is the classic liberal arts education. It rests on the belief that there are some set of basic skills and abilities necessary to participate fully in our democratic society. For example, such skills as literacy, numeracy and knowledge of history and civics.

I would further argue that American education is actually something of a hybrid system. We are attempting to provide a basic education to enable civic participation, but we allow some specialization (even at the secondary level; think vocational schools) in furtherance of career objectives. Now, whether we do this well or not is obviously an open question. Obviously, geography matters (when talking about primary/secondary education), as does socio-economic status.

Here I’d like to introduce another twist to the argument, which will also serve as a jumping off point for discussing sorting and signaling functions. One of the critiques of schooling (emanating from the political left; note here left means true left, not liberal Democrats) is that the whole notion of a common school is merely a myth. The argument is that there has never been a Golden Age of public schooling, that our schools have been used primarily as a means to assimilate immigrants and to inculcate certain American values, while stripping immigrant groups of their native cultures. In other words, public schools were not these great democratizing/leveling institutions, but rather a tool of social control.

Now, how does questioning this myth lead us to viewing higher education as performing sorting and signaling functions? I would like to argue that accepting the premise of sorting/signaling does not eliminate higher education’s actual education function. It would be foolish to do so.

However, in accepting this argument, we are more clearly able to explain some of the outcomes of higher education like pedigree bias. I would argue that due to inherent inequalities in primary and secondary education, much of what is believed to be merit based admissions to universities is by and large sorting. I would not go so far as to make a claim that high school GPA and/or SAT scores have no intrinsic merit. However, I do hold that given how well these indicators correlate with family incomes, they ought to be problematic.

Now, many post-secondary institutions take applicant background into consideration. Sometimes this may be directly through affirmative action programs. And it is certainly true that many/most institutions realize the inherent inequality of our education system and seek to redress some of that imbalance in their admissions policies. However, these types of admits provide a very small percentage of any institution’s students. (I will leave aside, for now, the question about affirmative action and how it may benefit higher socio-economic status minorities over their lower SES peers.)

Now, if my argument is valid, we can see how admission to post-graduate studies is compounding this inequity. Professional schools and graduate schools look not only to an applicant’s standardized test scores and GPA, but also to the “quality” of their undergraduate institution. So, if certain types of students (middle to upper income) tend to attend the most prestigious undergraduate institutions, then we can expect post-graduate training to be similarly skewed towards these same students.

That same argument can be made with respect to career opportunities after completing post-graduate studies, so I’ll not recount the steps.

The final argument, then, is this- our higher education system is merely perpetuating the existing social hierarchy in the larger society. It is not serving as a great democratizing force. Sure, some less fortunate students do pass through the gates of Harvard and then on to an elite graduate program. But, by and large the people going through our elite universities and our elite professional schools, are the people whose parents have also trodden that very same path.

What is actually being assessed is not the student’s merit, but his/her family background and connections. And this should be troubling to anyone who believes, as I do, in using our education system as a means to redress inherent inequalities in our society.

UPDATE: Ron Brownstein also touches on upward mobility.