Income and Top-Tier University Admissions

In a recent column on the connection between family income and top college admissions, David Leonhardt makes an interesting, but flawed, argument:

Several years ago, William Bowen, a former president of Princeton, and two other researchers found that top colleges gave no admissions advantages to low-income students, despite claims to the contrary. Children of alumni received an advantage. Minorities (except Asians) and athletes received an even bigger advantage. But all else equal, a low-income applicant was no more likely to get in than a high-income applicant with the same SAT score. It’s pretty hard to call that meritocracy.

In response, Greg Mankiw writes:

Maybe David is right, but to convince me, here is what I would like to see. Regress some measure of college success (such as GPA) on SAT scores and the student's family income. If David is right, then the coefficient on family income should be negative. That is, a lower-income student should do better in college, holding reported SAT score constant, because he managed to get that SAT score without all those extra benefits. This is a regression that some enterprising college admissions committee could easily do.

Mankiw goes on to post an email from an economist at The University of Western Ontario who ran just this kind of regression and found—unsurprisingly—that the coefficient on family income is positive, not negative. Thus, it seems top colleges should not expect students from lower-income backgrounds to receive higher grades, on average, than their middle-income peers with identical SAT scores.

These findings call into question at least one version of Leonhardt’s hypothesis. But I think the implications here are a bit more complicated, depending on how you view the concept of meritocracy.

If some “enterprising college admissions committee” did run Mankiw’s regression and found that the coefficient on family income is negative, what would this really mean? Putting aside the question of significance, it would tell us that a unit increase in annual family income (let’s say $1000) is generally associated with a certain percentage point decrease in GPA, holding SAT scores constant. In other words, among students with identical admissions test scores, the kids with higher family incomes tend to have lower college GPAs.

It seems obvious to me that this is not the case, as family income likely provides advantages beyond those captured by higher SAT scores.

There are lots of reasons why low-income students with good SAT scores would have lower college GPAs, on average, than their middle- or high-income peers. For one thing, higher family income is firmly linked to higher levels of parental education, which is a strong indicator of focus on education and education-based achievements. This matters in terms of performance, and it matters in terms of our expectations for low-income students who apply to top-tier colleges.

Family income is also likely to affect students after the admissions process. Kids whose parents make less money face all sorts of obstacles in college, like having to work longer hours to support their studies.

If you believe that a true meritocratic system should account for these disadvantages, it makes sense to argue that college admissions boards should give stronger consideration to an applicant’s household income. But this isn’t the same thing as saying students from lower-income backgrounds will perform better than other students with equivalent admissions test scores.

To be fair, I don’t think Mankiw is actually arguing that college admissions boards should completely discount income disadvantage. He’s simply pointing out a fundamental conceit in Leonhardt’s argument. Mankiw is correct that if an admission committee’s only metric is performance, it makes little sense to consider family income, except to the extent that this aids the committee in its selection of higher-achieving students. (One possible problem with Mankiw’s equation is that it assumes the relationship between GPA and SAT scores is identical at all income levels. Even for a performance-focused admissions committee, that assumption is pretty problematic.)

The bigger challenge in all of this is figuring out what we actually mean when we talk about meritocracy. Certainly, income provides unearned advantages, but so does innate ability. In a country that values both exceptional performance and equality of opportunity, college admissions boards need to strike some sort of balance.

While I agree with Leonhardt that income is one factor college admissions boards should take into account, I can’t seem to justify why it’s necessarily more important than other kinds of advantages.

-Jeremy

Strong Dollar = Red Herring

Christina Romer is a person I’m happy exists. She has the ability to explain complex ideas in ways that make sense on an everyday level. She recently wrote in the New York Times on our countries fascination with maintaining a “strong” dollar. At some point I’d like to dig more into a point Ms. Romer touches on, the tying of the Chinese renminbi to the dollar, but for the moment I’d like to focus on the fallacy that a “strong” dollar is always a good dollar.

The dollar, through its exchange rate, is a price. Market forces determine how much the dollar is worth and the government has relatively limited control over this process. Ms. Romer does a great job of explaining this process

The supply of dollars to the foreign exchange market comes from Americans who want to buy goods, services or assets from abroad. The demand for dollars comes from foreigners who want to buy from the United States.

Anything that increases the demand for dollars or reduces the supply drives up the dollar’s price. Anything that lowers the demand for dollars or raises the supply causes the dollar to weaken.

Consider two examples. Suppose American entrepreneurs create many products that foreigners want to buy, and start many companies they want to invest in. That will increase the demand for dollars and so cause the dollar’s price to rise. Such innovation will also make Americans want to buy more goods and assets in the United States — and fewer abroad. The supply of dollars to the foreign exchange market will fall, further strengthening the dollar. This example describes very well the conditions of the late 1990s — when the dollar was indeed strong.

Now suppose the United States runs a large budget deficit that causes domestic interest rates to rise. Higher American interest rates make both foreigners and Americans want to buy more American bonds and fewer foreign bonds. Thus the demand for dollars increases and the supply decreases. The price of the dollar will again rise.

This example describes conditions in the early 1980s, when President Ronald Reagan’s tax cuts and military buildup led to large deficits.

So what should this mean for our policy? Essentially trying to manipulate the dollar’s price is fool’s gold. It would be far more fruitful to a) understand what the price of the dollar is telling us about our economy and the impression others have of our economy and b) remove the win-lose connotation behind our descriptions of a “strong” and “weak” dollar

This second point seems simple but runs quite contrary to human nature. To essentially rebrand the dollar it would take a concerted effort to move away from the status quo. This effort is one that I think is completely worth it. Avoiding the hysteria that comes when we the see Euro valued at $1.41 is a worthwhile goal and would allow us to have a far more coherent monetary policy message towards the public (privately it seems pretty clear we already have these common sense policies in place). So to do our small part to help achieve this goal this blog promises to try and use the terms “expensive” and “inexpensive” when referring to the dollar in the future.


-Bo

Education and Federalism

Matt Ygelsias recently wrote on class, geography and education in response to E.D. Kain’s review of two education movies (one pro- and one anti-reform). My immediate response to this was slight indifference. Education is such a complex issues, and it seems like it might be impossible to get any sort of agreed upon solution, that I often tune out the debate. What eventually grabbed my attention in the Ygelsias article are the implications it has for the role that federalism has, and should have, when we think about education.

The majority of America’s children are not living in predominantly urban or poor districts. But much of the education reform movement is motivated by concern for this very group. The focus on these children makes sense. Achievement scores for poor, urban, districts are generally much lower than other school districts. It is completely natural to try to craft a solution to help those who are underserved.

What does strike me as problematic is that we often treat education in dichotomous terms. We seem to say that education for all districts ought to have this type of testing, or it ought not to. Standards should be the same in New York regardless of whether a district is in Ithaca or Buffalo. Partially, this arises because we like to think of ourselves as being fair. Why should we treat a poor district any differently than a rich district? I also believe that this is a side effect of the pro-reformers’ focus away from home life, which they cannot control, and onto what goes on when they have stewardship over students.

The problem with this of course is that all students are not the same. When I worked as an AmeriCorps Vista one of my jobs was to speak to graduating college students who were student teaching. There were a number of points that I tried to convey to these students, but the most relevant talking point for this discussion deals with class and language. Todd Risley, in his book Meaningful Differences in the Everyday Experience of Young American Children, writes about studies showing a major correlation between income/class and children’s verbal skills. For example, a child in a high-income household is likely to know 20,000 words by the age of six. A six-year-old in a household receiving welfare is likely to have a vocabulary about one-third of that size. The question that both pro- and anti-reform groups have to answer is: Should we treat both of these children the same way?

 

Jeremy and I recently had a talk about how federalism, and how I am generally mistrustful of it because of segregation, small groups’ propensity for pettiness, and the volatility it can create due to the fact that change is much easier on a smaller scale. One of the points he brought up, and that Ygelsias also brought into focus, was the positive effect it could have on education. Letting communities have more autonomy at the local level would seem to at least partially address difficulties state and national governments have in dealing with the vastly different needs of their communities.

 

Now this is not an argument against testing. Nor is it an argument against the federal government having any role in education. In fact, in some ways I think the federal government needs to be more involved, particularly in funding. It seems backwards to me that we leave the funding of schools up to communities that often have very different means, and then argue for nationalized standards.

 

In arguing for more federal funding I do not expect that the average amount spent per child for every state will go up. As states receive federal dollars, it is far more likely that they will cut their own funding, essentially maintaining the status quo. The groups that would hypothetically be impacted the most are those that at the moment do not spend what the government would provide per child. In that way, if a smaller community wanted to add more, it could. But every child would meet a minimum standard, with stipulations as to how competencies are tested and the role communities are to play in education.

 

This isn’t a proposal for a clear course of action. I think it’s important to note that there are differences between communities and that we shouldn’t be embarrassed by these differences. I am afraid of a movement towards homogeneity and the deifying of test scores in our education system. I much prefer our education system to that of China’s (a subject for another post) and I think that acknowledging the shades of gray in this debate will help us avoid some major pitfalls. As we move forward in the debate about how to shape our school systems, it seems incredibly useful to articulate the inefficiencies of a one-size-fits-all solution.

-Bo

As a teaser, I just wanted to let you know that Jeremy and I have asked our friend Veronica to guest blog on some education issues. She has graciously agreed, so expect to see some much more nuanced takes on this specific issue in the future. 

On Marriage Equality and the State

To follow up on Bo’s comments, I’d like to offer some of my own thoughts on marriage as a social institution and, more specifically, the extension of federal marriage benefits to same-sex couples.

As someone who’s deeply sympathetic the idea of marriage equality, but generally skeptical of the social necessity of state-sponsored matrimony, I find myself in a bit of an intellectual quandary. Arguments in favor of same-sex marriage often rest on the assumption that marriage is a social good that needs to be subsidized by the federal government. This position—which we may call the conservative argument in favor of gay marriage—is powerfully articulated by Andrew Sullivan in his book Virtually Normal. Sullivan writes:

Marriage provides an anchor, if an arbitrary and often weak one, in the maelstrom of sex and relationships to which we are all prone. It provides a mechanism for emotional stability and economic security. We rig the law in its favor not because we disparage all forms of relationships other than the nuclear family, but because we recognize that not to promote marriage would be to ask too much of human virtue . . . . A law institutionalizing gay marriage would merely reinforce a healthy trend. Burkean conservatives should warm to the idea.

This case is pretty compelling if you accept the premise that marriage generates positive externalities not clearly associated with other kinds of social arrangements. But that viewpoint doesn’t generally square with liberal orthodoxy on nontraditional households. To suggest that marriage deserves special status is to imply that alternatives to marriage—like cohabitation—are somehow less socially desirable, even if we choose not to "disparage" them. That’s not necessarily wrong, but it’s a difficult sell to someone who believes that all social arrangements are equally valid and the state has no place in promoting one kind of lifestyle over another. For a marriage skeptic like me, Sullivan’s argument simply isn’t very convincing.

But there are other arguments in favor of same-sex matrimony that don’t require such a positive reading of marriage as a social institution. In fact, some common points offered by advocates for marriage equality actually seem to provide a persuasive narrative for eliminating state-sponsored marriage. In response to traditionalist claims about the “stabilizing” effects of heterosexual marriage, some defenders of same-sex matrimony point to statistics on the steadily increasing rates of separation and divorce among heterosexual spouses. The obvious implication is that there is nothing particularly stabilizing about an arrangement into which people can so frivolously enter and exit.

If this is true, though, what does society gain by expanding such a weak institution? There are several possible answers to this question, but I’d like to consider the one that I find most legitimate. We may call this the principled pragmatist’s argument in favor of marriage equality.

It goes something like this: Without a realistic chance of eliminating federal marriage benefits, basic equality demands that we allow gay couples access to this institution. The key difference between heterosexual couples and homosexual couples is the ability to produce children who are biologically related to both parents. This distinction may have social relevance, but with an increasing number of married couples choosing not to have children, it no longer provides sufficient grounds upon which to limit marriage benefits to heterosexual couples.

For someone who’s not convinced of the social advantages of state-sponsored matrimony, this is much more appealing than the conservative position. But accepting the principled pragmatist’s argument naturally requires us to acknowledge other kinds of arguments that may make people a bit more uncomfortable.

If equality means that same-sex couples should be allowed to marry, and marriage has little to do with children who are biologically related to their parents, then why shouldn’t siblings also be allowed to marry? Genetic concerns no longer provide a compelling basis upon which to exclude sibling couples from the institution of marriage. (Even if this weren’t the case, there are important genetic concerns for other kinds of couples, but let's save this issue for another post.)

The argument that sibling couples have “alternatives” that are unavailable to gay couples also strikes me as unconvincing. Certainly, the inherent nature of homosexuality makes it different from incest in important ways. But the idea that siblings should be denied equality simply because they have other options is eerily reminiscent of claims made by anti-miscegenation advocates, who insisted that equality only necessitates a right to wed a person of the same race, rather than a person of the individual’s choosing. To suggest that siblings who love each other have “alternative” marital options is to apply similar logic: We all have an equal right to marry someone who is not an immediate family member, so we need not concern ourselves with the rights of sibling couples.

While most Americans today recognize that limiting a person’s ability to marry outside his or her race does not represent true equality, there seems to be enormous cognitive dissonance when this argument is applied to sibling relationships.

I understand that many are offended by allusions to incest in the context of a discussion over gay marriage. But I think it is incumbent upon those of us who see equality—rather than social stability—as the central issue in the debate over same-sex matrimony to give more serious consideration to other kinds of marriage equality arguments.

This is an area were reasonable people may disagree, and I admit that my thoughts here are not as developed as I’d like. For those of you who support or oppose gay marriage for different reasons—or just think my position is ridiculous—I’m curious to hear your opinions.

My goal here is to promote conversation, not to offer the final word on marriage equality.

-Jeremy

On Marriage

In an upcoming post, Jeremy is going to write about gay marriage, but before he does that we wanted to take a look at marriage in general, and examine if it should be an activity the state sponsors. For this discussion let’s assume the status quo, with the limitation that marriage can only be between a man and a woman. Should the state be involved in the licensing of marriages, and should the state provide benefits to married couples?

For proponents of state-sponsored marriage (and for sake the of this post, let’s assume the U.S. model, which has financial incentives for the married), I think the argument quickly becomes that the state should try to influence the actions of individuals to help them achieve a better life. In the case of marriage, the argument made by many proponents -- and the one that has the most resonance to me -- is that traditional nuclear family creates an environment that has a higher rate of success for children. We, as a society, want to encourage the success of children. Therefore we ought to increase the incentives, financially and by establishing social norms, towards the nuclear family.

The article “Marriage from a Child’s Perspective: How Does Family Structure Affect Children, and What Can We Do about It?” does a really good job at sorting out the settings in which children have the best outcome.  The first thing to note is that children from married, low-conflict, households have the best outcomes. What’s equally interesting to me are the caveats that the authors note. The quality of the parents’ marriage is hugely important, as is the fact that both parents be biologically related to the child (stepparents might actually be worse than single parent households depending on what study you read).  Co-habitation produces worse outcomes than marriage as well, but a major reason for this is noted in the study in that these relationships tend to be more fragile and are more likely to end within five years.

So if low-conflict marriage is clearly the best way for children to be raised, why wouldn’t it make sense for the state to incentivize this? Is there an argument against the government trying to establish circumstances in which people who want to pursue a nuclear family have the best chance of flourishing?

There are numerous reasons why someone might argue against the government being involved in marriage, but I’m going to present the two that appeal to me. First, the data is unclear.  If both parents are in a healthy marriage the same personality traits that make them successful partners seem likely to also carry over to child-raising. These personality traits are difficult to control for and incredibly important. Just as interesting to me is the fact that children with stepparents do notably worse than children without stepparents, even if both types of marriage can be considered healthy. This raises all sorts of questions about whether the type of parental relationship matters or how the parents themselves relate to the child. To claim that children raised in nuclear family households are more successful is one thing; to claim that they are more successful because they are raised in a nuclear family household is a much more difficult claim to prove, and one that the above authors are not making.

The second reason the government’s role in supporting marriages makes me uncomfortable is that it reinforces hetero-normative values that might be causing problems for other families from non-traditional backgrounds. While I don’t think this explains away the advantages of children in a low-conflict, two parent, household it certainly makes us ask important questions. Do we want to, as a society, engage in social engineering that necessarily results in those who are already outside the norm facing more disadvantages? If we are uncertain, and it seems impossible to ascertain, what the exact role family structure plays in the outcomes of a child wouldn’t it be better to provide financial support for all families generally?

I admit that this issue is somewhat personal for me.  I come from a single-parent household, and there were times when the pressures faced by me, but especially my mom, where exacerbated by hetero-normative expectations. The fact that I had a loving mother and father makes me incredibly lucky, but I have an inherent dislike of the government being involved in this. I hesitate though because granting the government a positive reading, that they want to help children by fostering a society in which the nuclear family can flourish, it seems that a lot of good might indeed come from marriage. Wouldn’t it be easier to admit that, yes, nuclear families have children with the best outcomes, we want that for our children, let’s figure out ways to support the flourishing of nuclear families? In the end however I think the truth is far too muddled for this type of action by the government. Yes children have better outcomes when the come from nuclear families, but as noted above, it is certainly unclear that these nuclear families are the cause of the better outcomes. There are real negatives in this too. Not only does in reinforce hetero-normative values which can have negative effects on non-traditional families, but it also establishes a precedent for the government to fundamentally involve itself in the personal lives of its citizens based on imperfect reasons. With these reasons in mind, and Jeremy might disagree with me in his later post, I tentatively argue against the state being involved with marriage.

-Bo

Nate Silver is a Liar, and I'm OK With That

One thing that fascinates me in everyday life is how we view large institutions. Specifically, I’m interested in the moral values that we assign to large corporations like Walmart, and how those values differ from the way we think about locally owned mom-and-pop stores. Since reaching my 20s, I’ve avoided illegally downloading music. I never felt right about it intellectually. Eventually my sense of guilt at downloading something that I know a) hurts someone whose work I’m enjoying and b) removes a financial incentive for that person to keep working overcomes any pleasure I gain from saving $10 on an album. However, I feel very little compunction about finding every loophole, and even watching my roommate stretch the truth, when dealing with my cable company.  A large part of this derives from the fact that it intuitively feels like my cable company would have no problem using the fact that it has a monopoly on my apartment building to offer me less than ideal service. In fact, I know that this is true from experience.

The reason I bring this history up is that recently Nate Silver wrote a column on how to beat airline fares using a “hidden-city” trick that caused some controversy over at the NYT. The basic premise of Silver’s idea was that airlines charge extra to fly somewhere if they have a monopoly on that city. It is often cheaper to fly much farther with a stopover in the city where the airline has a monopoly than to actually fly directly to that city. Chicago and Dallas/Ft. Worth are both examples of cities where it is traditionally much more expensive to fly into because United Airlines and American Airlines, respectively, have relative monopolies in these cities. So if you are flying from D.C. to Chicago, it can be cheaper to book a flight to Las Vegas and just not board the plane from Chicago to Las Vegas. Silver gets into the intricacies of how to successfully pull off this trick in his original column. However the real excitement comes from the comments and when Silver responds in the controversy in his column above.

The key point made by those who objected to this trick is that in the Contract of Carriage that one agrees to when purchasing a ticket (I never noticed that I agreed to this, but I have a bad habit of clicking as many boxes as I can, as quickly as possible, to end the pain of not looking at Cats That Look Like Ron Swanson), the passenger agrees not to use of this “hidden-city” trick. Silver responds by saying that because one does not have the option to negotiate a Contract of Carriage, this act of using the “hidden-city” trick is a way for customers to renegotiate with companies that have a relative monopoly on air travel (that is subsidized with government funds) on their Contract of Carriage. He furthers this point by noting that Contracts of Carriage are legally dubious and usually considered not binding.

My response to all of this is that I think Silver is wrong about the fact that this is a renegotiation. I think violating your Contract of Carriage is clearly breaking an agreement that was entered into willingly. If you do not like an airline’s policies you have the option to fly other airlines (Southwest has an awesomely lax Contract of Carriage) or not fly at all. Calling this a renegotiation is a way for Silver to avoid getting into tough questions like: Is it wrong to lie to a major company that is distorting a market, which also relies on taxpayer dollars? I think rather than tell ourselves we aren’t lying, it would be more useful to admit that the majority of us do not feel bad about lying to airlines on a topic like this because we have a pretty bad opinion about airlines. So until I see an argument that convinces me that airlines deserve more respect than my cable company, I am in the Nate Silver camp, with the caveat that this definitely makes me a liar.

-Bo

Some Thoughts on a Driving Tax

Since this blog was inspired as much by distance as anything else, it seems appropriate to begin with a look at the recent proposals to institute a driving tax.

There's been a fair amount of discussion lately about supplementing or replacing the current federal gas tax with a mileage tax that could, in theory, provide a more stable source of revenue for highway maintenance. The idea has already come under fire by some civil libertarians, who have raised concerns about the federal government potentially tracking private vehicles. I'd like to put these objections aside for the time being to make a basic point about the notion of replacing a gas tax with a mileage tax.

In my view, these taxes are fairly distinct. While revenue generated through either a gasoline tax or a mileage tax could be used to maintain the federal highway system, the two policies really have different aims.

The primary goal of a gas tax is to change people's behavior. By raising the price of gasoline, we're maing sure that individuals who pollute are responsible for the social cost of their decisions. These kinds of taxes are called Pigovian taxes, and they're designed to correct market failures. Revenues generated through Pigovian taxes are almost always less stable than those from other sources because we're trying to get people to respond negatively to the tax and purchase less of the item. This means less money for the government.

A mileage tax is not a Pigovian tax, but a user fee. This kind of tax is designed to ensure that people who drive on interstate highways are responsible for the cost of highway upkeep. However, it's important to note that taxing mileage doesn't directly encourage people to reduce their gasoline consumption. Someone who drives a hybrid would be taxed at the same rate as a someone who drives a similar vehicle with far less fuel economy. While a mileage tax would likely encourage people to drive less, it wouldn't necessarily encourage them to substitute away from inefficient vehicles. 

These may be fairly obvious points, but it seems like people are beginning to see these two taxes as interchangeable. They're really not. If we're trying to stabilize revenues for federal highway maintenance, the mileage tax makes an awful lot of sense. But that doesn't mean that we should stop trying to internalize the cost of pollution due to private gasoline consumption.

-Jeremy

Welcome!

Welcome to Proportioned Beliefs, a collaborative weblog focusing on American politics and policy issues, with occasional forays into popular culture. A quick background on who we are and what we’re attempting to do with this space. Bo and I recently earned our graduate degrees in public policy, but we now find ourselves moving on to different careers in different cities. Since we've spent much of our free time over the past two years talking about wonky ideas, we'd really like to continue some of these discussions over the Internet.

Our hope is that this blog will motivate us to keep talking, and provide a forum for others to join in the conversation. Discussions will likely center on policy issues that catch our attention or that seem to have some relevance to current affairs. We also have a few ideas for longer-term projects that we think may be interesting, but please keep in mind that this is a work in progress! Like all great experiments, ours runs the risk of degenerating into self-indulgent nonsense.

Ultimately, we're just trying to have fun with this blog, while continuing to talking about things that pique our interest. Enjoy!

-Jeremy