Predicting the future is a bit like singing karaoke: Many enthusiastically volunteer to do it in public thinking they’ll impress, but after their ineptitude inevitably reveals itself they just hope everyone will move on and forget how badly it went. When I chose behavioral economics as a dissertation topic, I heard constant predictions about its bright future from friends, advisers, and the news. Throughout the past decade or so, respected business and media outlets have consistently run bold claims like: “We are on the verge of a totally new way of thinking about economic rationality” (Psychology Today, 2009); “behavioral economics will shake marketing to the core” (Campaign Magazine, 2010); “behavioral economics can reinvent HR” (Deloitte, 2015); and others besides. Similar excitement has also been manifest in well-known podcasts, Malcolm Gladwell books, and TED talks.
Behavioral economics can be roughly described as the Venn-diagram intersection of social psychology and traditional economics; or, as the second coming, in the form of a new designer methodology, of traditional microeconomics following its lengthy eclipse by macroeconomics (and its handmaiden econometrics) over the past roughly 75 years. Its admirers believe that it combines the rigorous methodologies and consequential topics of traditional economics with the understandability and relatability of social psychology. To take an example, one behavioral economics research paper shows that consumers are significantly more likely to buy convertibles on warm days than on cold ones, even though the idiosyncrasies of weather on the day of a car’s purchase are irrelevant to both the owner’s enjoyment and its utility over the rest of its lifetime. Much of the well-known research in behavioral economics purports to show that humans are systematically irrational, and that the suboptimal quirks of the mind’s decision processes have meaningful effects on markets and economies writ large.
Besides the predictions about behavioral econ as an academic field, there has been a great deal of buzz about its role in government. In 2009, the legal scholar Cass Sunstein coauthored the bestselling book Nudge with economist Richard Thaler. Having described the theory and results of behavioral economics, Sunstein and Thaler make a case for paternalistic government policies that correct or take advantage of human irrationality for social good. Sunstein went on to head the Office of Information and Regulatory Affairs in Washington, where he pushed a host of nudge-style behavioral economics interventions. Around the same time, the UK and U.S. governments both formed high-profile teams assigned to apply behavioral economics research to the efficient functioning of government. These teams have been referred to colloquially as “nudge units” because of Thaler and Sunstein’s book.1
Academic papers arguing for government policies informed by behavioral economics have appeared at a fast and ever-quickening rate. The recently founded Behavioral Science and Policy Association has attracted several prestigious names to its board, including New York Times columnist David Brooks and Nobel Prize winner Daniel Kahneman, and has begun to publish its own journal advocating for and attempting to guide the implementation of behavioral economics ideas in government. In 2016 Sunstein asserted again that behavioral economics should and will have an important, growing role in governments worldwide. Citing “nudge units” in various countries, Sunstein wrote, “nations should certainly be encouraged to consider creating them. Their emergence and proliferation is a significant gain; there should be more of them.”
In short, behavioral economics has had a banner decade, and its prestige in academia, industry, and government grows apace. Predictions about the field’s promise still mount, and the excitement vibrates on—especially on the political Left. It is not hard to explain why.
First, because of the longstanding leftward lean of academia in general and social psychology in particular, most self-described behavioral economists are leftists or at least conventional liberals. Second, the irrationality that behavioral economics claims to uncover in typical decision-makers provides a justification for increasing government power by creating nudge-y laws and regulations to protect citizens from their own irrational selves. So for big-government progressives eager to increase the state’s regulatory power, behavioral economics furnishes what seems like a scientific justification for their favorite paternalistic policies.
The most intriguing reason for the excitement about behavioral economics on the Left, however, is that a handful of scholars have done research that (explicitly or implicitly) impugns political conservatives. Some such papers argue that conservatives are more biased than progressives or suffer from certain psychological pathologies more than their progressive counterparts.
Most of this anti-conservative research is from the social-psychological wing of behavioral economics rather than its economics wing, and explicit anti-conservative research has never constituted more than a handful of serious papers. Still, given how tenure and seniority work in universities, conservatives might understandably fear the (allegedly) scientifically justified disdain of their political enemies. Others might just find it uncomfortable—to put it gently—to be accused by academic peers of being insane in the membrane.
The numerous predictions of the past decade about the eventual triumph of behavioral economics in science and government have been made in this context: a field whose practitioners are mostly self-styled progressives, that provides scientific justification for progressive policies, and that even claims to prove that progressives’ political enemies are mentally deranged. For a leftist, it clearly would not take much nudging to sing this field’s praises and make lofty predictions about its eventual success.
But more recently, several developments have doused these glowing predictions. Most obviously, with the Republican electoral sweep in November, big changes are coming to the Washington bureaucracy. Regulatory agencies will be staffed with more conservatives, and they may also become smaller, more toothless, or both. There is a very good chance, too, that their leaders will be vastly less interested in behavioral economics and in using it to nudge the electorate.
But bureaucratic shifts are the least of the concerns that behavioral economics faces. Its real long-term problems are the fundamental issues with the science behind it. Taken together, these are grave enough to invalidate the handful of extant research papers that traduce conservatives. They are also serious enough to cast doubt on the efficacy of the nudge-style interventions in government that have so excited progressives.
One issue that behavioral economics faces is fraud. Every year or so, it seems, a much-lauded paper gets retracted because its data were discovered to be fraudulent. In late 2014, for example, Michael LaCour (then a Ph.D. student) was one of two authors on a research paper in Science entitled “When Contact Changes Minds.” The paper reported on a randomized, controlled trial of activist canvassing in which homosexual and heterosexual canvassers were asked to go door-to-door and have conversations with voters in California precincts that had supported Proposition 8 (a ballot measure banning gay marriage). Homosexual canvassers were asked to reveal to their interlocutors that they were homosexuals who wanted to get married, and heterosexuals were asked to reveal that they were acquainted with homosexuals who wanted to get married.
The paper reported striking results. After conversations about gay marriage that lasted about twenty minutes on average, voters’ support for gay marriage jumped up considerably. Support for gay marriage among voters who were visited by canvassers who were homosexual rather than heterosexual underwent a large, sustained increase that continued for almost a year.
LaCour’s results also fit perfectly with the liberal/progressive worldview. For one thing, the directional increase in support for gay marriage among all canvassed groups tracked with progressives’ common belief that they are on the “right side of history,” and that the arc of public opinion will inevitably bend toward what they believe is right. For another, the research told an inspiring story of a marginalized minority group gaining power over its uncertain destiny. BuzzFeed News ran a headline highlighting this inspirational side of the story: “Scientists Report Gay People Are the Best at Changing Minds on Marriage Equality.”
The study wasn’t only a happy affirmation of liberal opinion. It also disparaged conservatives. Consider: Its principal finding was that short conversations with homosexuals can have enormous and long-lasting effects on conservatives’ opinions about gay marriage. What kind of voters could be so completely swayed on important issues in the course of one introduction that lasted only minutes? Voters who had spent real time considering the facts of an issue, or years deciding on and living according to deeply held principles, might be mildly influenced but would not be completely persuaded by one short conversation with a stranger. If twenty minutes is enough to turn them around on a complicated issue, they are likely to be pathologically gullible or capricious. That the canvassing was only effective when conducted by homosexuals would imply that these conservatives were not only capricious, but insular and ignorant to the point that they had never met an avowed homosexual before the canvassers reached their doors.
Besides these highly improbable implications, there is the more obvious ideological slant of LaCour’s study manifest from its very first sentence, which explains opposition to gay marriage as the result strictly of prejudice. Neither LaCour nor his coauthor, or his reviewers or the journal editor, evidently thought that there could be any principled reason to oppose gay marriage, or any other reason at all besides prejudice. The piece oscillates between using the terms “opinion change” and “prejudice reduction.” But they only measured changes in opinions about gay marriage laws, and provided no measurements of prejudice itself (relevant if one believes that one can oppose gay marriage laws without being prejudiced against homosexuals).
The consonance of LaCour’s study with today’s left-wing orthodoxies likely helped it achieve publication in Science as well as receive excited media coverage in outlets like the New York Times, the Wall Street Journal, the Atlantic, and other important publications. But this extraordinary attention also led to its downfall. After the paper’s rise to prominence, some independent researchers found several irregularities that led them to question whether the paper was a fraud. It was: LaCour never conducted the study he described. Instead he created, as his innocent but highly malleable coauthor described it, “an incredible mountain of fabrications with the most baroque and ornate ornamentation.”
Fraud is a problem in every human endeavor, but academia lacks rigorous safeguards against it. Reviewers and journal editors typically regard their role as auditors of the purely scientific aspects of research: experimental design, statistical analyses, theoretical soundness, and likely impact. Peer reviewers usually take authors at their word: If they say they collected data, reviewers believe it, and a paper relying on fraudulent data can get published as long as it is plausible and its scholarship seems sound. Only a tiny fraction of published social science papers are ever revealed as frauds, but in the past few years, “data vigilantes” like Uri Simonsohn at Wharton have uncovered fraud in surprising places, and have developed tools to more effectively detect frauds in the future. Eventually, then, we ought to get less of it as the chances of being caught in a lie rise.
Although it is conceivable that there are more, if not many, fraudulent behavioral economics papers that remain undetected, it is unlikely that data fraud is very common. Besides, an even more serious and notorious problem in social science is the “replication crisis.” Scientific replication is simply this: Researchers write a paper describing a result and the methods they used to obtain it, and then other, independent researchers follow the described methods on their own and check whether they obtain the same results. If a scientific finding is true, it need be apodictic, meaning that we may expect any competent researcher to replicate it by following the same methods.
In social psychology (and other sciences, including economics and medicine), increased interest in replicating experimental findings has developed during the past few years. To the surprise and chagrin of many academics, a huge number of trusted experimental findings have not replicated when attempted by independent researchers. The quantity and importance of findings that have not replicated is so great that the problem is now referred to as a crisis. The crisis is deep and broad, cutting across disciplines and into long-established research programs, and it has no obvious or easy solutions.
One idea in social psychology that has recently been criticized as frequently unreplicable is “priming.” In priming experiments, subjects are exposed to simple stimuli, and then measured for behavioral outcomes related to the stimuli only through indirect psychological associations. For example, one well-known social psychology study attempts to show priming by exposing subjects to words related to the elderly, then surreptitiously measuring how fast they walk. The social priming theory being tested is that merely reading a word associated with elderly people will trigger mental associations with the elderly, and these triggered mental processes will manifest themselves in observable behavior.
The original paper was a big hit, and has been cited thousands of times in other research papers. However, numerous attempts to replicate the result have all failed. Given the number of failed replications, the prevailing consensus is that this effect does not exist, or if it does that it is negligibly small or has a very short half-life. There are numerous other papers on different kinds of priming, including the priming effects of different colors of text and those of pretending to smile, that have similarly failed to replicate.
Replication is usually performed on one study at a time. Occasionally, however, researchers stumble upon a methodological problem with an entire body of literature, casting doubt on years of seemingly meticulous research. This was the case with a recent methodological critique of the statistical tools used in functional MRI (fMRI) studies. A group of researchers found that literally thousands of papers published over many years had relied on a statistical method that was flawed, yielding false positive rates of 70 percent rather than the 5 percent rate that researchers had assumed. With false positive rates this high, it is to be expected that a huge proportion of the thousands of results established using this method will not replicate—because they are not true.
Researchers have occasionally used fMRI methods to study brain activity that correlates with different political affiliations and beliefs. Many have proffered claims about what these fMRI studies can reveal about conservatives and liberals. One Huffington Post article claims that these studies show that conservatives are more “squeamish” than liberals, because their brains apparently have stronger reactions to disgusting images. Another claim is that the conservative brain has a lower aptitude for detecting, tolerating, and dealing with uncertainty than the liberal brain. However, the recent discovery of the problems in fMRI analyses casts serious doubt on both of them.
The failure of behavioral economics studies, including those about priming, to replicate properly poses a serious problem. Since there is uncertainty about exactly which effects replicate and which don’t, there is now at least a little doubt about nearly all of the main ideas of behavioral economics, including those presented in Nudge. Scientifically, the field needs to spend years sorting through which effects are real and which are not. Politically, it becomes harder to argue for nudge-style behavioral economics interventions in government if we are not even certain that they work. If we are going to accept paternalism and the increased state power and expenditure that go with it, we should at least be confident that we will get some return on our investment.
Still other problems lurk beneath the surface. One serious potential problem with psychology that is rarely, if ever, discussed is that the assumptions underpinning the field could themselves lead to misleading or unfair conclusions about conservatives (or other groups). Consider the “status quo bias” as a case in point.
This bias is the tendency to resist change more often than warranted by ordinary cost-benefit analysis, for example by refusing to switch banks to get a better interest rate. The status quo bias certainly exists; each of us knows someone who is stubborn and pigheaded enough to resist nearly any change, however positive.
The problem is that it is rarely clear from a normative standpoint when change should be resisted and when it should be embraced. Maybe switching banks to get better rates is a no-brainer, but maybe the service is better at one’s current bank, maybe there are transaction costs, maybe there is uncertainty about whether contracts will be honored, maybe there are better loyalty rewards or conveniences at one’s current bank, maybe the difference in rates is too small to justify the time and bother of making a change, and so on. To say that the status quo bias is responsible for a decision is to implicitly pass judgment on how good or bad the status quo as a whole is compared to the alternative—to focus on interest rates in this hypothetical case is to isolate one element from a broader context, and that itself can easily be deemed irrational. Alas, one man’s status quo bias can be another’s rational choice.
Ever since Edmund Burke, conservatives have generally resisted significant changes to social institutions while liberals have generally embraced them. Thoughtful conservatives could give several good reasons for resisting change, including the Oakeshottian idea that it’s better to favor certain present laughter over unlikely future bliss, or any of the reasons that Burke gives in his Reflections. But leftists are not likely to regard these justifications for resisting societal change as valid. Progressives, excited to make rapid and often radical changes, tend to regard the status quo as ipso facto unacceptable and think that anyone who defends it (like conservatives) are evincing a status quo bias.
I have heard psychologists with progressive worldviews express the idea that all of political conservatism is simply a manifestation of the status quo bias—an assertion that is an opinion of political philosophy masquerading as psychological science. This masquerade is all too common and it is worthwhile to keep the two fields separate. Psychology is never very far removed from philosophy, but sometimes a scientist can be talented in one but not the other. Great scientific sophistication in psychology, economics, and statistics can be marred by philosophical fallacies such as the notion that defending the status quo must always involve some sort of a bias.
In Jonathan Baron’s decision theory textbook Thinking and Deciding, three references are listed in the index for “status quo bias,” including an entire section about it. Tellingly, there are no entries that refer to any kind of corresponding “change bias,” or tendency to embrace change more often than one should. One might expect that such a bias would exist, since it stands to reason that for every stubborn conservative who resists change too much there is an overenthusiastic liberal who embraces it too much. Could this asymmetry perhaps suggest a left-leaning bias in contemporary psychology?
There are other theoretically dubious biases that conservatives are accused of evincing. Take “threat bias,” which is the supposed tendency to perceive things too often as threats. Liberal psychologists have argued that conservative opposition to illegal immigration and support for long sentences for criminals both manifest this bias. Just as with status quo bias, however, there is no normative or analytical clarity about how much sensitivity to threats is too much sensitivity (one man’s threat is another man’s benignity). So was Neville Chamberlain right and Winston Churchill wrong after all? And as with status quo bias, one pretty much never sees psychological studies about a “complacency bias” or a tendency to view things as threats not often enough.
It is difficult for conservatives to argue that they don’t suffer from these biases, since doing so is not just a matter of crunching numbers or running experiments, but of arguing against established theories now embedded within the whole field of psychology. Behavioral economics, which relies on social psychology for much of its theory and ideas, will suffer in its pursuit of truth to the extent that the assumptions of psychology are dubious or ideologically biased.
Another problem in the social sciences is “construct validity.” In the physical sciences, it is a straightforward matter to agree on what is being studied and how to measure it. A kilogram of uranium (say) is easily identified as a kilogram of uranium by anyone with the right training. In the social sciences, however, research often requires an imaginative leap from an idea to its experimental implementation. To study a proposition like “people with open personalities are more likely to support open borders,” one cannot pour “personality openness” into a beaker, weigh it, and experiment with it. One must create a “construct,” which amounts to something quite else that can be taken to be a reasonable proxy for the object of study. In psychology, to measure something like openness, researchers frequently choose a battery of questions that have been validated in prior literature.
The need to create constructs to study psychological ideas introduces a whole new potential source of error and bias into behavioral economics research. In a recent City Journal article, John Tierney pointed out some of the poor constructs that have been used in psychology studies about conservatives:
[O]ne study explored ethical decision making by asking people whether they would formally support a female colleague’s complaint of sexual harassment. There was no way to know if the complaint was justified, but anyone who didn’t automatically side with the woman was put in the unethical category. Another study asked people whether they believed that “in the long run, hard work usually brings a better life”—and then classified a yes answer as a “rationalization of inequality.” Another study asked people if they agreed that “the Earth has plenty of natural resources if we just learn how to develop them”—a view held by many experts in resource economics, but the psychologists pathologized it as a “denial of environmental realities.”
Sometimes, the connection between constructs and the object of study can be weak enough that two nearly opposite interpretations can apply to the same data set. For example, one paper examined correlations between male upper-body strength and political conservatism. Among wealthy men, upper-body strength had a relatively strong correlation with conservative attitudes about redistribution. (The paper explained this correlation through an appeal to evolutionary theory.) At first, this may seem like a compliment to male conservatives, suggesting that they are virile and strong. However, upper-body strength was not measured directly, but rather through the “construct” or proxy of bicep circumference. Prominent statistician Andrew Gelman has referred to this study several times on his blog as a study about “fat arms,” implying that wealthy conservatives are not strong but fat, or perhaps misshapen. This creative interpretation has the remarkable power, without altering the original paper one whit, to turn a scientifically justified compliment to an unmitigated insult.
One could also creatively interpret constructs to turn insults into compliments. One study attempted to examine the relationship between the psychological intolerance of deviance and politically conservative attitudes about law and order. The construct the authors used to study “intolerance of deviance” was the extent to which subjects regarded incorrectly drawn triangles, squares, or circles as representative of the classes of triangles, squares, or circles, respectively. As reported by Quartz:
In a series of studies involving 2,100 US adults, psychologists . . . asked participants to rate perfect and imperfect shapes on a scale of triangle-ness, circle-ness, square-ness. In some experiments the researchers showed shapes with dents or lines missing, while other studies had lumpy or oddly stretched shapes.
Researchers also collected information about participants’ political beliefs, asking them to report on how liberal or conservative they are. They asked participants questions about how severely criminals should be punished, whether homeless people should receive public aid, and how they felt about both.
The team found that participants who self-identify as conservative were also less forgiving of non-perfect shapes. They were more likely to punish criminals and be more morally outraged by them. Participants who were more sensitive to imperfect shapes were significantly less likely to want to provide aid to the homeless, unemployed, and uninsured sick.
Corresponding author Tyler Okimoto writes in an email to Quartz, “We believe this reflects a general tendency for conservatives to judge deviant targets as “different from the norm” compared to liberals, be they criminal offenders, recreational drug users, the homeless, same-sex couples . . . or even geometric figures. And as a result, conservatives are more likely than liberals to believe these groups should face harsher punishment or less aid.”
The authors’ interpretation represents an enormous and unjustified ideological leap from the collected data. If conservatives had conducted these same studies, they could have just as easily proposed interpretations that painted liberals in a bad light instead. For example, they could have claimed that the data indicated that conservatives are better at geometry than liberals. The definitions of triangle, square, and circle have been clearly and strictly defined ever since Euclid. Euclid himself, had he taken these surveys, would have answered in harmony with the very strictest conservatives, not allowing any “lumpy” or dented shapes to be counted. And yet liberals, ignoring the clear rules of geometry, counted these non-shapes as shapes more often than did conservatives. Could these results be taken as proof that liberals are less intelligent than conservatives, or that they are anti-geometry and therefore anti-science? The possibility is not considered; the researchers go straight from the data to the slur that conservatives are psychologically intolerant of deviance. To go directly from judgments about triangles to conclusions that conservatives are “less forgiving” than liberals is a wild example of ideologically motivated overreach, but that sort of thing is also wildly common today in some social science domains.
Still other scientific problems facing behavioral economics are too numerous and complex to fully delineate here. These problems include inflation of estimated effect sizes in research literature, “Type S” errors in which hypotheses are falsified by detecting an effect of the sign opposite the true effect, and a host of problems with the frequentist null hypothesis significance-testing paradigm that remains dominant.2
Other political challenges also abide. If many nudge-style interventions (like social priming) do not end up replicating, behavioral economists in government will have a harder time justifying their paychecks. If social psychologists and behavioral economists continue to push ideas that unfairly malign conservatives, they could get fired and watch as their programs are gutted every time a Republican takes office. The study and implementation of behavioral economics interventions could also be prohibitively expensive in practice because of the complexity involved in both.
With all of these challenges, behavioral economics faces a hard road ahead in the worlds of both science and politics. The field will have to heal itself, by rooting out fraud, ensuring that effects replicate properly, cleaning up biased theoretical assumptions, and ensuring that constructs more closely match intended objects of study. In the course of this healing process, it should become a more ideologically balanced field. It should also take into account serious counterarguments about the proper role of paternalistic behavioral economics in government—an issue that Sunstein himself has admitted is a serious one—and thereby decrease its overreaching paternalistic ambitions.
For conservatives concerned about anti-conservative research bias and how to respond to it, several intellectually sound and responsible approaches are available. The first is denial. When liberal scientists claim to have proven that conservatives are of unsound mind, conservatives can point out the nonzero fraud rates, close-to-zero replication rates, theory problems, construct problems, statistical issues, and practical issues that behavioral economics faces, and how together they suggest that none of the extant anti-conservative research is certain to be accurate.
Denial is fine, but many conservatives seek to score points of their own in the ideological battle for the heart of social science. They want to go on the offensive. So perhaps the first study could be on “the tendency to do social psychological studies that malign one’s political enemies” as the first liberal pathology. They might add a study of the pathological inability to classify basic geometric shapes. But this sort of thing is on balance a bad idea. Social science doesn’t need a never-ending war of name-calling and escalating accusations. It needs a genuine scientific orientation to the subject matter. If erstwhile pioneers in behavioral economics (before the label even existed) such as Herbert Simon and Kenneth Arrow managed to achieve this, it stands to reason that future researchers can, too.
A bit more humility all around would be useful as well. One recent estimate holds that according to the current Diagnostic and Statistical Manual (DSM), almost half of the population (more than 46 percent) will suffer from a diagnosable mental disorder at least once in their lives. Professional psychiatrists like Theodore Dalrymple have objected to this conclusion on the basis that it is too high, but I think it is too low. I have met many people all around the world—liberal, conservative, and everything else—and I have yet to meet one who was a perfect specimen of ideally functioning psychology. Each of us has some sort of quirk, and for what it is worth, I am joined in this opinion by Moby Dick’s Ishmael.
Raised a Presbyterian, Ishmael viewed Presbyterianism as the plain truth and other religious traditions as lunacy. But he came to see that Presbyterians and pagans, though different, were both groups that, as is common to all humans, suffered from psychological pathologies:
I say, we good Presbyterian Christians should be charitable in these things, and not fancy ourselves so vastly superior to other mortals, pagans and what not, because of their half-crazy conceits on these subjects. There was Queequeg, now, certainly entertaining the most absurd notions about Yojo and his Ramadan;—but what of that? Queequeg thought he knew what he was about, I suppose; he seemed to be content; and there let him rest. All our arguing with him would not avail; let him be, I say: and Heaven have mercy on us all—Presbyterians and Pagans alike—for we are all somehow dreadfully cracked about the head, and sadly need mending.
Melville’s insight is twice valuable. Yes, we are all imperfect, but just as important, “all our arguing” with those outside of our ideological group “would not avail.” An implication of this, it seems to me, is that a devotion to (social) science and truth should motivate us to make sober and honest inquiries into our own mental issues as well as those of our ideological foes.
As behavioral economics progresses, it may actually uncover conservative biases in ways that are free from the problems outlined above. But if so it should also identify biases typical of liberals. The future of behavioral economics in government will depend on bigger battles about the size of government and the regulatory state, and social scientists will not be the final arbiters of that argument. It behooves conservatives and liberals to be more charitable toward their political foes, and in that direction lies the interest of social science itself.
1Not everyone applauded the book, and that included some who in no way could be described as conservative. See Steven Teles, “Nudge, or Shove?” The American Interest (January/February 2015).
2Empirical researchers are trained from Stats 101 onward to calculate “p-values,” which provide a measure of how unlikely one is to find particular results given a “null hypothesis” that a phenomenon being studied does not exist. For decades, statisticians have been pointing out serious problems with p-values and this method of gathering evidence by falsifying null hypotheses. A classic paper on the problems of p-values is available here, and a much more recent statement on p-values cautioning against their misuse from the American Statistical Association is here. In many fields, finding a p-value that is sufficiently close to 0 is all that one needs to clear the hurdle of being worthy of publication. However, the papers cited above indicate that this is not a reasonable criterion for scientific evidence. Essentially, serious statisticians argue that a variety of respected empirical fields are misusing what should be a narrowly applicable statistical method and are allowing untruths to enter the research literature because of it.