December 5, 2019 2:45PM

What’s the Matter with Significance?

By Peter Van Doren and David Kemp

Discussions about statistical significance are not usually found in newspapers, but the Associated Press recently had such a discussion about the results of a clinical trial involving a heart drug. Statistical significance refers to whether a study finds a “real” effect or whether any differences measured are a result of chance. For example, in the case of the heart drug study, the authors attempt to measure whether the drug reduces patients’ mortality by comparing the mortality of patients on the drug to people not on the drug. The statistical significance reflects the authors’ confidence that the difference (reduced mortality) they find is not a fluke. As the article correctly states, “Significance is reflected in a calculation that produces something called a p-value. Usually, if this produces a p-value of less than 0.05, the study findings are considered significant. If not, the study has failed the test [i.e., the findings cannot be differentiated from random chance].”

The heart drug study had a p-value of .059, meaning that the study’s authors are 94.1 percent confident that the apparent lower mortality they found for patients on the drug than for those not on the drug is real. By the standard .05 p-value (95 percent confidence level) criterion, the study’s findings are not considered statistically significant.

In all scientific research (including clinical trials), scientists must make a tradeoff between the likelihood of accepting a false finding because it does reach statistical significance even though it is a statistical fluke (a false positive or Type I error) or rejecting a real finding because a study doesn’t reach statistical significance (a false negative or Type II error). Reducing the probability of Type I errors inherently increases the probability of Type II errors, and vice versa. In medicine, this tradeoff is between the possibility of ignoring a beneficial medical finding that does not reach statistical significance and endorsing a fruitless or harmful medical finding that does reach statistical significance.

The lead investigator of the heart drug study believes that by rejecting their findings because they don’t reach a p-value of .05 their study falls in the former category: “the drug in fact produced a real benefit and…a larger or longer-lasting study could have reached statistical significance.” But the article also refers to a study of another heart drug that “found a significant treatment effect for patients born in August but not July, obviously just a random fluctuation.”

What should be done about these tradeoffs? The article correctly states that the traditional cutoff of .05 is arbitrary and should be abolished. Instead, studies should report the p-value and other accompanying evidence to allow the reader to decide how to use the results of scientific and medical studies.

These are old issues. Van Doren discussed them over 10 years ago in a review of The Cult of Statistical Significance by Stephen Ziliak and Deirdre McCloskey in the Cato Journal. The AP article echoes many of Ziliak and McCloskey’s points and the relevance of statistical significance and other seemingly scientific questions to public policy and decision-making remains. Many health and safety decisions are delegated to bureaucracies that allegedly use scientific methods to decide what products and practices to allow on the market. In reality, values enter into these decisions in a variety of ways, including questions about how large sample sizes should be, the costs and benefits of decisions, and what level of statistical significance is accepted.

Policy debates often ignore the value questions inherent even in scientific research and fail to recognize that people with different values will come to their own conclusions based on the information available to them. Regulators and researchers should gather and disseminate this information without injecting their own values or rejecting findings based on an arbitrary level of statistical significance.

December 5, 2019 12:25PM

Give the Gift of P. J. O’Rourke

The Cato Institute offers lots of great Christmas gifts -- Pocket Constitutions (also a good gift for Bill of Rights Day!), books, apparel, even Cato-branded Lands' End merchandise. But I have my own holiday recommendations that I've made before.

Media Name: PJ3.jpg

I decided one year to give a young colleague a post-graduate course in political science and economics — P. J. O’Rourke’s books Parliament of Whores and Eat the Rich. So I went to my local Barnes & Noble to search for them. Not in Current Affairs. Not in Economics. No separate section called Politics. I decided to try Borders (RIP). But first — to avoid yet more driving around — I went online to see if my local Borders stores had them in stock. Sure enough, they did, in a couple of stores just blocks from the Cato Institute. Checking to see where in the store I would find them, I discovered that they would both be shelved under “Humor–Humorous Writing.” Oh, right, I thought, they’re not books on economics or current affairs, they’re humor.

Yes, P.J. is one of the funniest writers around. But what people often miss when they talk about his humor is what a good reporter and what an insightful analyst he is. Parliament of Whores is a very funny book, but it’s also a very perceptive analysis of politics in a modern mixed-economy democracy. And if you read Eat the Rich, you’ll learn more about how countries get rich — and why they don’t — than in a whole year of econ at most colleges. In fact, I’ve decided that the best answer to the question “What’s the best book to start learning economics?” is Eat the Rich.

On page 1, P. J. starts with the right question: “Why do some places prosper and thrive while others just suck?” Supply-and-demand curves are all well and good, but what we really want to know is how not to be mired in poverty. He writes that he tried returning to his college economics texts but quickly remembered why he hated them at the time–though he does attempt, for instance, to explain comparative advantage in terms of John Grisham and Courtney Love. Instead he decided to visit economically successful and unsuccessful societies and try to figure out what makes them work or not work. So he headed off to Sweden, Hong Kong, Albania, Cuba, Tanzania, Russia, China, and Wall Street.

In Tanzania he gapes at the magnificent natural beauty and the appalling human poverty. Why is Tanzania so poor? he asks people, and he gets a variety of answers. One answer, he notes, is that Tanzania is actually not poor by the standards of human history; it has a life expectancy about that of the United States in 1920, which is a lot better than humans in 1720, or 1220, or 20. But, he finally concludes, the real answer is the collective “ujamaa” policies pursued by the sainted post-colonial leader Julius Nyerere. The answer is “ujaama—they planned it. They planned it, and we paid for it. Rich countries underwrote Tanzanian economic idiocy.”

From Tanzania P. J. moves on to Hong Kong, where he finds “the best contemporary example of laissez-faire....The British colonial government turned Hong Kong into an economic miracle by doing nothing.”

You could do worse than to take a semester-long course on political economy where the texts are Eat the Rich and Parliament of Whores. So, bookstore owners, leave them in the Humorous Writing section for sure, but also put copies in the Economics, Politics, and Current Affairs sections.

Still time to buy them for Christmas and educate all your family and friends while they think they're just being entertained!

December 5, 2019 9:51AM

Despite More Staff, CBP Says “No Resources” To Process Asylum Applicants At Ports

Customs and Border Protection (CBP) has taken a series of unprecedented actions to limit the ability of immigrants to request asylum in the United States. But among its earliest and most consequential decisions was to cap the number of migrants who it would process for asylum at ports of entry. This policy clearly violates federal law. More importantly, it forces asylum seekers to remain homeless in squalid and desperate conditions in dangerous Mexico border cities, leading many to cross illegally.

The American Immigration Council and Al Otro Lado have challenged the policy (the government calls it “metering”) in court, and the government has argued that it lacks the resources to process undocumented migrants who arrive at ports to request asylum. In other words, it is violating the law because it simply physically cannot follow it. The challengers argue that this is a mere pretext for keeping out asylum seekers, and administration officials comments seem to confirm that this is the intention.

The data support their contention that the Trump administration has artificially reduced the capacity of ports without any reduction in resources for ports. In response to a Freedom of Information Act request, CBP provided me the number of CBP Office of Field Operation (OFO) officers permanently assigned to southwest border ports of entry. CBP-OFO processes documented and undocumented travelers who apply to enter at ports.

Figure 1 compares the annual number of “inadmissible aliens” along the southwest border—generally those without proper documents proving pre-approval to enter, many of whom want to request asylum—to the number of CBP-OFO permanently assigned officers. As it shows, the Obama administration in 2016 processed 82,106 more undocumented migrants than in 2012 and 2013—more than double the earlier amount—despite 97 fewer CBP-OFO officers. The Trump administration added officers every year—increases of 692, or 11 percent—yet it cut port processing.

But the annual figures significantly understate the reductions in port processing under this administration. In the month of October 2016, the Obama administration processed 20,524 undocumented migrants at ports. In October 2019, the Trump administration processed just 9,733, 53 percent less than the peak processing month, as Figure 2 shows. The average CBP-OFO officer at a port of entry went from processing 3.1 undocumented migrants per month to just 1.4 per month, a 56 percent cut.

Nor is it a matter of undocumented aliens appearing in one region that cannot handle them. Figure 3 breaks down the numbers by CBP-OFO field office. In all four regions, the numbers have fallen in absolute terms: 57 percent in San Diego, 38 percent in Tucson, 44 percent in El Paso, and 56 percent in Laredo. Only El Paso sector—with the fewest migrants—had fewer officers in 2019 than 2016 (a 2 percent staffing drop). Laredo oversaw the processing of 4,884 fewer undocumented migrants in October 2019 as October 2016, despite 235 more officers (10 percent increase). San Diego allowed in 3,537 fewer, despite 120 more officers (6 percent increase). Tucson processing fell 1,428, despite 143 more officers (18 percent increase).

Figure 4 depicts the number of undocumented migrants processed at ports per officer for each field office. Again, relative to the officers available, the Trump administration is processing far fewer migrants than the Obama administration in all four field offices. Compared to October 2016, the average CBP-OFO permanently assigned officer processed 59 percent fewer in San Diego, 47 percent fewer in Tucson, 43 percent fewer in El Paso, and 60 percent fewer in Laredo than in October 2019. The average officer in Laredo went from handling 3.6 cases in October 2016 to 1.4 in October 2019.

The per-agent figures also highlight metering’s absurd justification. The average CBP-OFO permanently assigned officer processed just 1.4 migrants in the entire month of October 2019. The resource excuse strains the credulity of the court and the public.

These numbers actually understate the issue because CBP also temporarily transferred an unknown number of officers to deal with “overflow” operations. While we don’t know the numbers of officers, CBP revealed that it transferred $6.7 million in funds to deal with “overflow” operations at southwest ports of entry in 2016. In 2017 and 2018, it transferred $11.3 and $8.4 million, respectively, and in 2019, the number rose to $28.1 million by July 22—a fourfold increase over 2016. In other words, there were far more temporary resources available in 2019 as well.

Ultimately, the purpose of metering is make immigrants so miserable that they give up and go home, but while it undoubtedly does make them suffer all sorts of atrocities, it also drives many of them to cross illegally. This makes illegal immigration worse and results in unnecessary, dangerous crossings. A man and his daughter who were turned away at a port of entry ended up drowning in the Rio Grande River in July.

This fact also further frustrates the “resources” excuse because this summer CBP found ways to process vastly greater quantities of undocumented migrants who crossed illegally between ports of entry. At the same time, it claimed a total inability to increase the numbers at ports of entry. DHS continues to evade responsibility for causing illegal immigration and its deadly consequences.

December 5, 2019 9:09AM

Happy Repeal Day!

Today is a great day for freedom. On this day in 1933, the 21st Amendment was ratified, thus repealing Prohibition. My former colleague Brandon Arnold wrote about it a few years ago:

Prohibition isn’t a subject that should be studied by historians alone, as this failed experiment continues to have a significant impact on our nation.

Groups like the Women’s Christian Temperance Union, a key force in the passage of Prohibition, survive to this day and continue to insist that Prohibition was a success and advocate for dry laws.

Prohibition-era state laws, many of which are still on the books today, created government-protected monopolies for alcohol distributors. These laws have survived for three-quarters of a century because of powerful, rent-seeking interest groups, despite the fact that they significantly raise costs and limit consumer options. And because of these distribution laws, it is illegal for millions of Americans to have wine shipped directly to their door.

The website RepealDay.org urges celebrations of the "return to the rich traditions of craft fermentation and distillation, the legitimacy of the American bartender as a contributor to the culinary arts, and the responsible enjoyment of alcohol as a sacred social custom." It's easy! You don't have to hold a party. Just go to a bar or liquor store and have a drink.

RepealDay.org says that "No other holiday celebrates the laws that guarantee our rights." I think that's going too far. Constitution Day and Bill of Rights Day do exactly that. And in my view, so does Independence Day. But that's quibbling. Today we celebrate the repeal of a bad law. A toast to that!

Cato celebrated the 75th anniversary of repeal with this policy forum featuring Michael Lerner, author of Dry Manhattan: Prohibition in New York City; Glen Whitman, author of Strange Brew: Alcohol and Government Monopoly; Asheesh Agarwal, Former Assistant Director of the Federal Trade Commission’s Office of Policy Planning; and Radley Balko, Senior Editor, Reason.

December 4, 2019 5:41PM

It’s Easier to Criticize Pain Patients Than to Criticize Drug Prohibition

Last week the Washington Post featured an article in the "Outlook" section by Dr. James D. Hudson, a pain management specialist, lamenting Americans’ “Dangerous Fear of Pain,” arguing that the efforts by doctors to make their patients “pain free” has largely contributed to the overdose crisis.

Public policy towards the overdose crisis is panic-driven., Often lacking a strong basis in evidence and fueled by the media, it marks a return to the opiophobia of the 1950s and 1960s. This has led to comments about people in pain like this one by former U.S. Attorney General Jeff Sessions at a Tampa press conference in February 2018:

"Sometimes you just need to take two Bufferin or something and go to bed.”

As doctors are pressured by law enforcement and policymakers to curtail the treatment of pain, it seems a market has developed for articles by pain management doctors urging patients to “suck it up” and learn to live with pain. Dr. Hudson responded to the market demand. While not going to extremes like Attorney General Sessions, he still encourages people to learn to live with pain. At one point he writes:

In many ways, learning to live with chronic pain is like learning to live next to the El in Chicago. This aboveground transit system is old and loud. It roars past homes and businesses. Nearby residents feel the vibrations; it can be so noisy that they pause their conversations while the train goes by. Yet those who think a lower rent is worth the annoyance report that the sound soon ceases to be disruptive. You get used to it.

Well, this doctor disagrees. I believe that part of the physician’s mission is to end pain and suffering, not just treat illnesses and injuries.

Unfortunately, this return to a pre-modern approach to pain management will do nothing to stop the overdose rate, on a steady exponential increase since at least the late 1970s—well before the creation of OxyContin. And while the obsession over the prescription of pain medicine to patients in pain continues, policymakers, Dr. Hudson, and the media continue to ignore the obvious: the overwhelming majority of overdose deaths are from heroin and illicit fentanyl obtained by non-medical users—most of whom were never patients—in the dangerous black market made possible by drug prohibition.

This moved me to write a letter to the editor of the Washington Post, which was published here today.

December 4, 2019 10:28AM

DeVos Proposes “Government Corporation” Handle Student Aid. Color Me Dubious

Gold bars

The U.S. Department of Education was not created to be a giant lending institution. But that is what it has become, overseeing nearly $1.5 trillion in federal student loans. At a meeting of college financial aid administrators in maybe-symbolic Reno, NV, Secretary of Education Betsy DeVos proposed to change that by turning the department’s student aid offices into an independent “government corporation.” It would have its own governing board that would, she suggested, be freed of political meddling and motivated to “deliver world-class service to students and their families.”

Color me dubious.

For sure, it is hard to imagine an independent agency running less efficiently than the U.S. Department of Education, though that is mainly because the bigger the organization, the greater the red tape. I am not aware of evidence that given its size and that it is a government institution, ED is a particularly atrocious steward of student loans. The fact is, as DeVos pointed out in her speech, it is Congress, with approval of the president, that has created the inscrutable profusion of loans, grants, and loan repayment plans—not to mention tax-favored savings accounts and Work Study—that has made federal student aid a tangled web only a professional financial adviser could love. Probably no one could make that into a well-oiled machine.

DeVos laid out few specifics for her proposal, but if an independent Federal Student Aid (FSA) agency had to follow all the rules and qualifications of the current aid programs—and Congress put them in there for a reason—it is hard to imagine the system getting much more efficient no matter who oversees it. With its only mission running aid programs, an independent FSA might be more efficient in watching over schools and loan servicers, but probably not by a lot. And DeVos suggested it would not just be a narrower focus that would make it work better, but greater incentives: a need to “work to secure its financial strength and sustainability.”

I don’t see it. If its job were to run student aid programs as they exist now—to deliver entitlement money right from the Treasury—the financial sustainability of the new FSA would be assured. It would also have no real competitors to push it to do better—private lending is less than a tenth of the volume of generous federal loans—and as a government entity it would presumably have no profit motive. In other words, it would have no incentives to be more efficient or consumer-friendly than ED.

It could also become dangerous. While DeVos said a new FSA should not “look like another CFPB,” one wonders what would keep it from targeting specific types of schools, or servicers, for scrutiny. True, its governors would not be facing popular election, so they might not have as much incentive to demonize for-profit college or unpopular servicers as politicians would, but everyone likes to be applauded for doing the popular thing.

Until this proposal is fleshed out more it should not be dismissed. But from what we have so far, it’s hard to be very optimistic. Indeed, DeVos said that even with the new agency the federal government would continue to “monopolize student lending.” As long as that is the case, we will be nowhere near where we need to go.

December 3, 2019 4:28PM

Most Central Americans Still Evade Mexican Migrant Crackdown

Mexico’s government has touted to President Trump its efforts to reduce migration of Central Americans through its territory to the United States and took credit for the drop in arrests of migrants at the U.S.-Mexico border since June. Yet Mexican immigration enforcement fails to explain the drop in immigrants reaching the United States for three reasons: 1) most who try are still making it, 2) the effectiveness of Mexico’s enforcement is slightly worse than the median month since October 2014, and 3) the drop has occurred mainly among families and unaccompanied children, while Mexican enforcement has mainly targeted adults traveling alone.

Figure 1 shows the number of arrests of immigrants from Central America’s Northern Triangle—Guatemala, Honduras, and El Salvador—by U.S. Border Patrol or Mexican authorities. The data reveal that Mexican enforcement is not only not unprecedented, but it is actually less effective than it was on numerous occasions in recent years. In September 2019, 61 percent of arrests occurred in the United States, indicating that most Central Americans evaded the crackdown. This is worse than the median month (59 percent) since October 2014, meaning Central Americans were less likely to reach the U.S. border in a majority of other months.

Figure 2 compares the likelihood of reaching the United States by the type of migrant. As it shows, the crackdown was more effective at stopping single adults than children with parents or unaccompanied children. More than 70 percent of unaccompanied children and children with parents made it to the United States in September 2019, compared to just 46 percent of single adults. Despite this focus on single adults, it was families and children whose numbers declined the most, falling 85 percent since May. Single adults have come down comparatively less: 64 percent. Many other months saw stricter enforcement.

Given these facts, it is more likely that other factors explain the absolute decline in arrests by the U.S. Border Patrol. The most important would be the inability of Central Americans to make asylum claims in the United States and the return of Central Americans into Mexican border cities under the “Remain in Mexico” policies. These actions—which have exposed numerous immigrants to crime, homelessness, and death—have likely deterred others from trying to seek safety in the United States.

Children and families are more likely to fear extended homelessness in Mexico, which explains why they were disproportionately affected by these policies. Single adults are also more likely to attempt to evade detection, so they are less influenced by changes in asylum policy. The United States has better alternatives to deal with this crisis than abandoning its asylum laws and treating asylum seekers like criminals. It should increase legal immigration and process asylum seekers at ports of entry.