201804

April 10, 2018 12:13PM

Why All the Hype About NAEP?

Here we go again. The 2017 round of the Nation’s Report Card was released today. The results shouldn’t surprise anyone – they are almost entirely flat at the national level. However, that doesn’t stop educators and education reformers from spinning the results to fit whatever agendas they might have. Those who defend previous reforms claim that computer-based testing must be to blame for stagnant performance – and that students today are “relatively poorer” than they were in the past. On the other hand, groups calling for additional reform claim that the NAEP results should startle Americans.

We should all settle down. Here are 3 reasons why:

  1. Test scores tell us very little about success.

Education scholars such as University of Arkansas’s Jay P. Greene have been talking about the weaknesses of standardized test scores for a long time. Specifically, Greene frequently points out that at least 10 rigorous school choice studies indicate disconnects between effects on test scores and effects on long-term outcomes such as attainment and earnings. In fact, Diane Ravitch recently praised Greene for shining a light on this issue. And it’s not every day that Ravitch and Greene agree on something.

Research reviews indicate that Greene is on to something. For example, a recent review of the academic evidence on the subject finds that “there is a weak relationship between impacts on test scores and later attainment outcomes.”

Similarly, I have started to review the causal private school choice literature that indicates divergences between effects on test scores and other long-run outcomes such as student criminality, effort, and happiness. As shown in the table below, I’ve found 11 disconnects in the literature since 2001. For example, the sample of students from the state-mandated evaluation of the Milwaukee voucher program saw no statistically significant improvements in reading test scores after the fourth year. However, those same students were more likely to enroll in a four-year college and less likely to engage in criminal activity later on in life. In other words, putting too much weight on standardized tests – like NAEP – could compromise the character development necessary for real lifelong success.

Studies with Disconnects Between Effects on Test Scores and Long-Run Outcomes

 

Media Name: disconnect_pic.png

2. Even if test scores were strong measures of long-run success, changes in NAEP score averages alone do not tell us much about changes in nationwide performance.

Let’s assume test scores mattered. Even then, an uptick in NAEP scores wouldn’t necessarily tell us that overall performance improved. People like former U.S. education secretary Arne Duncan claim that the student population “is relatively poorer and considerably more diverse” than in the past. He may have a point. For example, single-parent households have nearly tripled since 1960. However, many scholars also point out that inflation-adjusted per capita income has nearly doubled since the 70’s, while the share of citizens with college degrees surged over the same period.

Since the magnitude – and the direction – of the student population’s relative advantage changes over time, we cannot confidently determine whether NAEP performance has actually gotten better or worse.

3. Even if test scores were strong measures of success, and we could somehow know that changes in NAEP scores reflected changes in nationwide performance, we would have no idea what policies affected those changes.

Now let’s also assume that – somehow – student advantage was held perfectly constant over time. Even then, we would have no idea what specific policies affected the changes in student performance. States that have different education policies have different populations and laws that could also affect their average standardized test scores.

Significant gains were made in Florida, for example. In fact, Florida was ranked the best-performing state for 4th grade math and reading scores after the Urban Institute controlled for several student background characteristics. But why? As a supporter of private school choice, I could spin these results in my favor by pointing to the fact that Florida has over 100,000 students participating in their privately funded private school choice program. But the truth is that the experts currently have no idea if the school choice programs in Florida have anything to do with the gains.

None of us should get this hyped up over standardized test scores. After all, these crude NAEP score averages unfortunately cannot tell us much of anything absent rigorous empirical analysis. Anyone claiming otherwise needs to slow down and reassess the situation.

April 10, 2018 11:05AM

No Major Lessons from New National Test Scores

Another set of national exam results—the National Assessment of Educational Progress (NAEP)—is upon us, and much will likely be made of them. But in the aggregate, what the new scores show is just that things haven’t changed much over the last couple of years, and only as captured by this particular test. Burrowing down and comparing states, subgroups of kids, and smaller jurisdictions that have implemented different policies, spent more or less, and experienced numerous other things, might suggest some avenues for further exploration, but the only conclusion we can state with any confidence is that nothing happened—not Common Core, not school choice, not the Every Student Succeeds Act (ESSA)—that appears to have seismically altered NAEP outcomes.

That may be just fine: Americans have increasingly and broadly rejected standardized tests scores as the end-all-and-be-all of education, culminating in the ESSA, which in late 2015 replaced the No Child Left Behind Act and its obsession with testing. And as we are increasingly learning, tests scores may have little connection to other outcomes like high school completion, and neither really addresses whether kids are learning desirable moral values, or creative thinking, or the myriad other big things parents want for their children.

That important preface offered, what are the highlights, such as they are, in the latest NAEP?

First, that overall scores for 4th and 8th graders—no high school kids this time around—were statistically flat between 2015 and 2017 for all but 8th grade reading, which went up two scale points, from 265 to 267. (Note, they were at 268 in 2013, so stagnation persisted over four years). Within those numbers, top scorers tended to see scores rise, and lower scorers scores decrease, while lower-income students tended to see stagnation or slight dips.  Private schools—actually, only Catholic schools are reported—also saw general stagnation.

Media Name: naep.jpg

Maybe there are some standout states? For math, only Florida saw scores rise in both 4th and 8th grade, hitting a new high in the state for 4th grade, but not yet returning to the state’s 8th grade peak in 2013. In reading no state saw increases in 4th grade, but Florida was joined by nine other states with increasing 8th grade scores. So Florida may be doing more right than any other state, but what that is NAEP can’t tell us. It has been a very choice-friendly place, but also fond of test-driven accountability systems and even requiring kids to pick “majors” in high school. Heck, maybe it just does the best standardized testing prep, which may explain a lot of score increases in the past. (Corey DeAngelis will have more on what NAEP can't tell us soon!)

Finally, 21 districts have new NAEP scores. Not surprisingly given the national results, most saw little change between 2015 and 2017. San Diego, though, saw statistically significant increases in 4th grade math and reading, so maybe that city bears watching.

Two years of new data is not very much, and a whole lot influences standardized test scores. And frankly, such scores tell us only a fraction—maybe a very tiny one—of what parents want to know about education. So let’s hope we don’t lose our collective heads over these scores, and do not see a new policy obsession with standardized testing outcomes like we experienced under No Child Left Behind. Instead, let’s let individual educators and parents—who value so many things—choose what they think is most important in education.

April 10, 2018 10:07AM

Blame Mexico First

There are many aggravating tropes that keep reemerging in the debate over immigration policy but one of the worst is that every problem with the U.S. immigration system is the result of the supposedly perfidious Mexico.  Changes in Mexican law and policy certainly have an impact on immigration to the United States but it is not true that our laws would operate wonderfully even if foreign governments had policies to support them.  Those who blame Mexico should, at the minimum, get their stories straight. 

In many versions of this tale, the Mexican government is hypocritical because its immigration laws are strict yet it complains about laws like SB 1070 in Arizona and the deportation of Mexican citizens from the United States.  Talk radio show host Rush Limbaugh famously used this rhetorical tactic when sarcastically (maybe?) proposing a series of immigration reforms that mirrored the worst of Mexican immigration law in 2007.          

The Mexican government fiercely criticized the passage of Arizona’s SB 1070 in 2010, a bill that forced state and local police to enforce federal immigration laws.  When then-President of Mexico Felipe Calderon visited the White House and intended to complain about the law directly to President Obama, Representative Ted Poe (R-TX) said, “I wonder if they'll discuss whether or not Calderon supports his own country's immigration policy.”  Both Limbaugh and Poe rightly criticized Mexico’s famously restrictive, self-destructive, and hypocritical immigration policy.

Partly in response to the American criticism, Mexico gradually reformed its immigration laws beginning in 2008.  In that year, Mexico reduced the punishment for illegal entry to a maximum fine of 5,000 pesos, down from a potential ten-year prison sentence.  They also created a temporary agricultural guest worker visa program for agricultural laborers from Guatemala and Belize working in Mexico’s southern states.  In 2010, the Mexican government stated that illegal immigrants would not have to fear immigration enforcement when reporting human rights violations or receiving medical treatment. 

The Mexican government did not stop there.  The Mexican Congress passed a Migratory Act in 2011 that went into effect on November 1, 2012.  This law replaced the Mexican General Law of Population that was the source of their restrictive immigration laws.  Among other things, the new Mexican immigration law allowed immigrants and migrants equal access to Mexican courts, reduced the power of local police to enforce immigration law, reformed the humanitarian admissions system, simplified entrance and reduced residency requirement by, in part, creating a points system, and created a three-day regional visitor's visa for people from neighboring countries.  In other words, Mexico liberalized and expanded its legal immigration system.  Although Mexican reforms did not go far enough, they were a significant step away from protectionism toward a more liberal immigration regime. 

Shortly after the Migratory Act of 2011 went into effect, the number of Unaccompanied Alien Children (UAC) apprehended by Border Patrol skyrocketed (Figure 1).  The change in Mexican immigration law allowed Central Americans to travel to the U.S. border in greater numbers which, combined with the worsening economic and crime problems in Central America, helped exacerbate a surge of non-Mexican (OTM) illegal immigrants and asylum seekers who were overwhelmingly from south of Mexico’s border (Figure 2).  Other non-Mexican legal changes like the Central America-4 Border Control Agreement, that created a visa-less Central America among El Salvador, Guatemala, Honduras, and Nicaragua, also made it cheaper for migrants or UAC from those countries to make it to Mexico and then to the United States.    

 

Figure 1

Southwest Border Monthly Border Patrol Apprehensions of Unaccompanied Alien Children

Media Name: nowrastehpngmexico.png

Source: Customs and Border Protection.

 

Figure 2

Other than Mexican and Mexican Southwest Border Apprehensions

Media Name: nowrastehpngmexico2.png

 

Source: Customs and Border Protection.

As the surge of UACs heated up over the summer of 2014, the American government started pushing Mexico to strictly enforce its own immigration laws before the migrants got the American border.  All of a sudden, Mexico’s immigration laws had become too lax and American politicians and commentators were furious.  The hyperbole from the U.S. government rose to such levels that Marine Corps General John Kelly, then-SOUTHCOM chief, said, “[m]any argue these threats are not existential and do not challenge our national security. I disagree.”  The Soviet Union would have been an existential threat if it attacked the United States during the Cold War but it strained General Kelly’s credibility to claim that a surge of UAC and other migrants, equal to a small percentage of legal admissions that year, was a threat to the country’s existence.

During the 2016 presidential campaign, then-candidate Donald Trump blamed the Mexican government for America’s illegal immigrant population and frequently said that he will build a wall that Mexico will pay for.  The Mexican hypocrisy of defending illegal immigrants in the United States while having a strict immigration policy on its own southern border rankled Trump and many of his supporters despite recent changes to Mexican immigration law.      

Recently, President Trump criticized the Mexican government for not enforcing its immigration laws well enough in order to stop the caravan of about 1,000 Central Americans traveling to the U.S. border to ask for asylum.  As soon as the Mexican government largely broke up the caravan, President Trump quickly reversed course and praised Mexican laws for being stronger than American laws.

Mexico has serious problems.  Although it has made some strides in recent years, low economic and individual freedom have produced endemic corruption.  The U.S.-subsidized war on drugs has created a level of violence that occasionally reaches civil-war-level death tolls.  Regardless of those problems, Mexico should cease to be the blame for the failures of American immigration policy.  It is easy for American politicians to blame foreign governments rather than take responsibility for a sclerotic American immigration system that Congress has failed to substantially reform in almost three decades.  Recognition that Mexico has serious problems is at odds with the standard political received-wisdom that we should outsource much of our immigration enforcement policy to the very same government with so many problems.  Better American laws that allow more legal immigration through a cheaper, simpler, and better-enforced system will do a lot more to bring order to the border than relying on the Mexican government.  The policies of foreign government affect immigrant flows to the United States and no reasonable person could claim otherwise but the American political tradition of criticizing Mexican immigration law for either being too strict or not strict enough based on the U.S. news cycle should stop.  

April 9, 2018 3:52PM

CBO Report: Spending Blasts Off

The Congressional Budget Office has released new projections for the federal budget.

If Congress makes no reforms, federal spending is expected to rise from $4.14 trillion this year to $7.05 trillion by 2028, as shown in the chart. The rise represents an annual average growth rate of 5.5 percent, substantially higher than the 4.0 percent expected nominal growth rate of the economy.

The spending increases are expected to push up federal deficits from $804 billion this year to $1.53 trillion by 2028, while jacking up federal debt held by the public from $15.7 trillion this year to $28.7 trillion by 2028.

President Trump wants new spending on a NASA moon mission, but spending on existing programs is already blasting off for unknown fiscal frontiers.

Media Name: rocket.png

April 9, 2018 1:43PM

Federal Spending Rescission

Worried that their spending spree in the recent omnibus bill will suppress conservative turnout at the polls this November, Republicans are now considering a “rescission” package. The package of spending cuts—being designed by the White House—could be passed in Congress with simple majorities in both chambers.

The omnibus increased discretionary spending 13 percent in a single year, with large increases in both defense and domestic (nondefense) spending. The bill passed with majorities of Republicans in both House and Senate, and was signed by President Trump.

Some conservative commentators have suggested that Trump and the GOP did not want to increase domestic spending that much, but were pushed into it by the Democrats. Putting a spending rescission package—with cuts to low-value domestic programs—up for a vote would be a nice test of that theory, as it would not need Democratic votes for passage.

Trump proposed many cuts in his 2018 and 2019 federal budgets. Here are some that would be good candidates to include in a rescission package:

  • Ending the Community Development Block Grant.
  • Ending the Economic Development Administration.
  • Ending Essential Air Service subsidies.
  • Ending the USDA’s rural subsidies.
  • Ending various energy industry subsidies.
  • Ending the Low Income Energy Assistance Program.
  • Ending the Community Services Block Grant.
  • Ending the Weatherization Assistance Program.
  • Ending funding for the NEA, NEH, and CPB.
  • Cutting housing rental subsidies.
  • Cutting job-training subsidies.
  • Cutting farm subsidies.
  • Cutting foreign aid subsidies.
  • Cutting federal retirement benefits.

You can read about the merits of most of these cuts at DownsizingGovernment.org.

April 9, 2018 12:23PM

Michigan’s Attempt to Criminalize Speech in Support of Jury Independence

On November 24, 2015, Keith Wood stood on a public sidewalk in front of a courthouse in Big Rapids, Michigan, and distributed pamphlets published by the Fully Informed Jury Association (“FIJA”). These pamphlets discussed the history of the jury trial as an ancient and sacred safeguard of liberty, and the jury’s crucial role as the voice of the conscience of the community. It reminded jurors that they have the authority to engage in so-called "jury nullification" — that is, to refuse to return a guilty verdict when doing so would be manifestly unjust, even if legal guilt was proven.

In other words, Mr. Wood was engaged in speech at the very core of the First Amendment’s aegis: classic political advocacy (peacefully distributing pamphlets) in the quintessential public forum (the sidewalk in front of a courthouse) on a matter of public concern more ancient than Magna Carta, and at the heart of Anglo-Saxon law (the rights, duties, and independence of citizen jurors). But Mr. Wood’s advocacy nevertheless led to his arrest and conviction for violating a Michigan "jury tampering" statute that makes it a crime for anyone to "willfully attempt[] to influence the decision of a juror in any case by argument or persuasion, other than as part of the proceedings in open court in the trial of the case." Incredibly, neither the trial court nor the first-level appellate court thought there were any free-speech concerns with Mr. Wood’s conviction.

The Cato Institute has therefore filed a brief with the Michigan Court of Appeals, urging it to reverse Mr. Wood’s conviction. Under binding Supreme Court precedent, "[g]overnment regulation of speech is content based if a law applies to particular speech because of the topic discussed or the idea or message expressed." Reed v. Town of Gilbert, 135 S. Ct. 2218, 2227 (2015). That is exactly the case here; the statute specifically targets "argument or persuasion" on a specific subject matter — jury decision-making. The statute is therefore subject to strict scrutiny, which it cannot possibly satisfy. Although states certainly have a compelling interest in protecting the integrity of jury trials, this statute sweeps far beyond this legitimate purpose and implicates massive volumes of speech entitled to the highest degree of First Amendment protection. Moreover, the state has no interest — compelling or otherwise — in preventing Mr. Wood from discussing the history of jury independence, including the jury’s right to engage in "nullification."

Mr. Wood’s conviction is all the more troubling because the jury trial itself is dwindling to the point of a practical nullity — 97% of federal criminal convictions today are obtained through plea bargaining, not trials. We have, in effect, traded the transparency, accountability, and legitimacy that arises from public jury trials for the simplicity and efficiency of a prosecutor-driven conviction machine. There is no panacea for the jury’s diminishing role in our criminal justice system, but the least that states can do is not exacerbate the situation by singling out for punishment protected speech meant to inform jurors of their rights, obligations, and historical role as the conscience of the community.

April 9, 2018 11:13AM

Equal Pay Day Should Be in January

April 10th is Equal Pay Day according to the National Committee for Equal Pay. The holiday “symbolizes how far into the year women must work to earn what men earned in the previous year.”

Unfortunately, Equal Pay Day is a holiday created by people that don’t understand economics. If they did, Americans would celebrate the holiday in January.

Equal Pay Day is based on the gender pay gap, which compares the median woman’s wages to the median man’s wages. The resulting 18-percent-plus difference is often portrayed as attributable to gender-based discrimination.

But the gender pay gap is a flawed measure: comparing median men’s and women’s wages doesn't tell you anything useful. In order to make a fair comparison, researchers must compare male and female workers that have similar characteristics beyond being the middle worker in the income distribution.

Important characteristics include age, education, years of experience, job title, employer, and location. A recent Glassdoor study controlled for these characteristics and the gender pay gap fell from nearly twenty-five percent to around five percent.

Of course, if the remaining five percent gap represents discrimination that is a problem. But the Glassdoor study didn’t control for all relevant characteristics, and it’s likely the adjusted gender pay gap would fall a bit further if it did.

Americans should observe Equal Pay Day part-way through January, assuming the Glassdoor study is correct. Observing Equal Pay Day in January would paint a more accurate picture of the state of gender pay equity.