Tag Archives: mobility

Education Matters, But Direct Anti-Poverty and Inequality-Reduction Efforts Matter More

I once began a K-12 education talk by putting the following two questions on a screen.

1. What is the single policy change that would most improve the quality of K-12 education?
2. What is the single policy change that would most reduce the opportunity gap between low-income and high-income students?

I asked audience members to, by a show of hands, indicate which question spoke to them more.  They had three choices:

A) Question 1
B) Question 2
C) Doesn’t matter, since both question 1 and question 2 have the same answer

Stop and think for a second about which choice would have prompted you to raise your hand.

If you would have selected choice C, you would have been joined by about 90 percent of the audience at my talk.  I expected that result.  In a culture in which politicians routinely say things like “education is the closest thing to magic we have here in America” and cite low graduation rates in low-income areas as evidence of our education system’s failures, that view is unsurprising.

It’s also completely wrong.  The overwhelming evidence that choice C is incorrect falls into at least five primary buckets:

1) There are large gaps in test score performance in the United States before students enter kindergarten. The graph shown below, from the Economic Policy Institute, documents the extent of these gaps (there are gaps in various cognitive and noncognitive skills as well), and as Sean Reardon has shown, there is evidence that they close during the school year, only to reopen during the summer months.  The gaps have declined in size since the late 1990s, but they are, in Reardon’s words, “still huge.”

EPI Kindergarten.png

Inequitable access to preschool for low-income students is definitely part of the problem here, but gaps are apparent in infancy and probably due mostly to differences in housing, nutrition, medical care, exposure to environmental hazards, stress, and various other factors.

2) Decades of research into the causes of the gap in test scores between low-income and high-income students in the United States has consistently found a limited contribution from school-based factors. In the US, variations in school quality seem to explain no more than 33% of the discrepancies in test score performance; this number, which has been around since 1966, considers the influence of a student’s classmates to be a school-based factor (it arguably isn’t) and thus seems to be a conservative upper bound. Most studies put the school-based contribution to what is commonly called the “achievement gap” closer to 20%, with about 60% attributable to “student and family background characteristics [which] likely pertain to income/poverty” and the other 20% unexplained.

3) Economic success in this country is less common for low-income students who are successful in school than for high-income students who are unsuccessful in school. The graph below, made using data from the Pew Economic Mobility Project, compares the distribution of adult economic outcomes for children born into different quintiles of the income distribution with different levels of educational attainment.  If education were the prime determinant of opportunity, we’d expect educational attainment to determine these adult economic outcomes.  Yet the data show that children born into the top twenty percent who fail to graduate college typically fare better economically than children born into the bottom twenty percent who earn their college degrees.  In fact, the born-into-privilege non-graduates are 2.5 times as likely to end up in the top twenty percent as adults as are the born-poor college graduates.

Mobility - Pew

4) The test scores of students in the United States relative to the test scores of students around the world aren’t all that different than what students’ self-reports of their socioeconomic status would predict. The Programme for International Student Assessment (PISA) has an “index of economic, social, and cultural status” which incorporates family wealth, parents’ educational attainment, and more.  There is a gap in test score performance between students who score high on this index and students who score relatively low on it in every country in the world.  The size of the gap varies by country, as does the median test score, but there is a strong correlation overall between students’ socioeconomic status and their performance on standardized tests.  The first graph below, in which each data point relates the average socioeconomic index score for a decile of a particular OECD country’s students to that decile’s average performance on PISA’s math test, depicts this relationship.

OECD Test Scores - All.png

As the next two graphs show, test score performance for the bottom socioeconomic decile in the United States falls right on the OECD bottom-decile trend line, and while U.S. test scores for the second decile are a little below the OECD trend (as are U.S. scores for the next few deciles), socioeconomic status seems to explain American students’ performance on international tests pretty well overall.

OECD Test Scores - Bottom Decile.png

OECD Test Scores - Second Decile.png

5) The distribution of educational attainment in the United States has improved significantly over the past twenty-five years without significantly improving students’ eventual economic outcomes. While people with more education tend to have lower poverty rates than people with less education, giving people more education neither creates quality jobs nor eliminates bad ones, as Matt Bruenig has explained.  A more educated population (see the first graph below), therefore, just tends to shift the education levels required by certain jobs upwards: jobs that used to require only a high school degree might now require a college degree, for example.  The “cruel game of musical chairs in the U.S. labor market” (as Marshall Steinbaum and Austin Clemens have called it) that results is likely part of why poverty rates at every level of educational attainment increased between 1991 and 2014, as shown in the second graph below.

Bruenig1.png

Source: Matt Bruenig

Bruenig2.png

Source: Matt Bruenig

Bruenig’s analysis lacks a counterfactual – the overall poverty rate may well have increased if educational attainment hadn’t improved, rather than staying constant – but it’s a clear illustration of the problem with primarily education-focused anti-poverty initiatives.

None of this evidence changes the fact that education is very important.  It just underscores that direct efforts to reduce poverty and inequality – efforts that put more money in the pockets of low-income people and provide them with important benefits like health care – are most important if our goal is to boost opportunities for low-income students.

6 Comments

Filed under Education, Labor, Poverty and the Justice System

Good Policy Is the Goal. Compromise Should Not Be.

Matt Bruenig just wrote an excellent series of posts dismantling a misguided “Consensus Plan for Reducing Poverty and Restoring the American Dream” from the American Enterprise Institute and the Brookings Institution.  Bruenig’s posts explain why the plan’s emphasis on education, work, and marriage will not accomplish its goals (I’ve made similar points about education and family structure before).  While it’s important to note that education and work (and family strength and stability, which are critically different from family structure) have value – improving education and the availability of good jobs can boost economic mobility – the evidence is clear that we will not equalize opportunities for more than an unacceptably small subset of kids until we both reduce inequality and make sure kids’ basic needs are met.

Richard Reeves, a researcher I really like who participated in drafting this unfortunate “consensus plan,” describes it as a triumph of realism over purism.  In doing so, he draws a false equivalency between what he calls “purists on both political extremes: those on the right who simply see government as the problem, and fantasize about sweeping away vast swaths of institutional architecture and funding, and those on the left who imagine that simply taking money from some and giving it to others will cure society’s ills.”

“Liberals” (or, in the parlance of the report, “progressives”) and “conservatives” are the labels DC insiders typically use to categorize people on one or the other of these false extremes, as shown below.

LiberalConservativeBut the idea that these are two equivalently absurd “sides,” and that the best course of action is thus to compromise by meeting in the “middle,” is unfortunately a major impediment to good policymaking.  It is harmful primarily because it fails to capture how certain views and proposals are more ethical and evidence-based than others.

For example, if our goal is to reduce poverty and boost the opportunities of poor children, evidence shows that the “purists” Reeves describes on the “left” have a much more legitimate claim than those on the “right.”  Government programs like Medicaid, SNAP (formerly food stamps), and the Earned Income Tax Credit, for example, work very well on these fronts, as do direct cash transfers and more robust social insurance systems around the world.  Redistribution may not “cure [all of] society’s ills,” but it definitely works as intended in most cases, while gutting government programs, especially during times of economic hardship, doesn’t.  It is simply incorrect to suggest otherwise, but the categorization scheme above implies that each “side” has an equally legitimate perspective.

Consider how similar reasoning could be applied to current presidential debates on immigration.  Donald Trump’s platform (build a wall across the border, end birthright citizenship, and don’t let any poor people into the country, among other crazy ideas) could represent the perspective on one “side” of the political divide, while Bernie Sanders’ plan to bring 11 million people out of the shadows could represent the other.  Both Trump and Sanders say they want to put “the needs of working people first – [over those of] wealthy globetrotting donors” (Trump’s words).  The AEI/Brookings brand of “realism” could result in the adoption of a decent chunk of the Trump immigration agenda; it clearly isn’t an approach that makes for desirable policy.

Reeves is right that there are a “diversity of views” among those on each side of this uninformative partisan divide, and the AEI/Brookings team correctly notes that nobody “has a monopoly on the truth” – even Donald Trump occasionally has a good idea and even smart, principled politicians like Bernie Sanders sometimes get things wrong.  Yet a better political categorization scheme would explicitly note that Sanders’ policy positions are far superior to Trump’s on the two criteria that matter most: ethical considerations and the degree to which proposed policy ideas are supported by available evidence.  The tool below does so.

Political Tool.003

The x-axis is an “ethics axis” and requires us to think through John Rawls’ veil of ignorance.  As I’ve explained previously:

“Privilege-defending” viewpoints and policies that ignore the veil of ignorance – those that mainly consider the ideas, desires, and needs of people already in power – fall on the left side of this axis.  “Power-balancing” viewpoints and policies developed after reflection about the veil of ignorance – those that more ethically think through the concerns and needs of less-privileged people – fall on the right.  The vertical or “accuracy axis” of the tool orients us to the facts; it plots views according to the degree to which a combination of sound theory and empirical evidence informs them.

The ideal policy, developed with consideration of the veil of ignorance and using the most accurate interpretation of the facts, sits in the upper right hand corner.  This tool thus provides several advantages over…the traditional Left-Right spectrum.  First, it forces us to think about what matters; we cannot plot opinions on this tool without ethical and intellectual analysis.  Second, the tool captures that objectively good policy (policy in the upper right hand corner) is more desirable than the “center” of opposing viewpoints.  Third, it gives us a common framework to discuss policy ideas with people with different perspectives, orienting our conversation to two pillars – truth and justice – instead of normalizing disagreement as inevitable.

Elevating “bipartisanism,” “compromise,” and “realism” as goals might help a group come to a consensus wherein each “side” gets some things it wants.  It does not often result in good policy platforms, however, and the Brookings/AEI plan is a case in point.  If we want final products that are truly ethical and evidence-based, we need to reject compromise for compromise’s sake and start recognizing that some viewpoints and proposals are more legitimate than others.

2 Comments

Filed under Poverty and the Justice System, US Political System

Everything You Need to Know About Inequality

Jared Bernstein and I just published a comprehensive PowerPoint presentation on inequality in the United States (available for download here). This presentation is first and foremost intended as a resource. Part 1 of the presentation documents the increase in inequality over the past 35 years; the trend is evident from every major data set and income definition. Inequality deniers have fortunately become a rapidly-dwindling breed, but should you encounter one, Part 1 should help set the record straight.

Part 2 discusses why inequality matters. As we summarize on our slides, inequality “reduces opportunities, undermines the democratic process, distributes growth unevenly, and may even have negative macroeconomic effects.” We provide the mechanisms behind these reasons and evidence documenting them as well. The slide notes contain a more in-depth look at these issues, as they do throughout the presentation, for anyone interested.

Fortunately, increasing inequality is not inevitable; it is “a problem that better policies can directly address.” Part 3 of our presentation explores some of these policies, measures that can begin to address inequality’s causes and counter its effects. While our slides in this section are in no way exhaustive, they’re a good starting point for the type of comprehensive social justice agenda I’ve mentioned previously.

I believe the section on equality of opportunity (or lack thereof) in Part 2 is particularly relevant for education stakeholders, as it highlights the importance of this comprehensive agenda (as opposed to a narrower, education-only policy focus). Consider the following chart, the extended version of our featured graph from slide 26:

Source: Pew Economic Mobility Project

Source: Pew Economic Mobility Project

This chart uses data from the Pew Economic Mobility Project to compare adult economic outcomes for two different types of students: students who grow up in the bottom income quintile but manage to graduate college (“poor college grads”) and students who grow up in the top income quintile but don’t make it through college (“wealthy non-graduates”). Wealthy non-graduates (49%) are almost twice as likely as poor college graduates (27%) to end up in one of the top two income quintiles as adults.

As we discuss in our presentation, inequality presents a variety of education-specific obstacles for low-income students, and addressing these obstacles (and thus facilitating college completion) is an important part of our agenda. At the same time, the less favorable distribution for low-income college graduates shown above provides a critical reminder of the importance of a comprehensive social justice agenda. If we truly want to provide low-income students with opportunities equivalent to those of their higher-income peers, we need strategies that address the needs of low-income families, strategies that address some of the direct effects of growing up disadvantaged. That means we must get the economy back to full employment. It means we need a strengthened safety net, higher labor standards, and a reversal of the decline in unionization. It also means we need better-regulated markets and improved fiscal policy. A push for quality education can’t have its intended effect unless it’s part of this larger agenda.

Jared and I hope that this presentation, by documenting inequality’s rise and consequences, provides a compelling basis for these claims. While the historically high level of inequality represents a serious threat to fundamental American values – opportunity, democracy, and broadly shared prosperity – we believe we can combat this problem if we acknowledge it and work together on solutions.

5 Comments

Filed under Education, Poverty and the Justice System, US Political System