Logical Fallacies in the Time of Pandemics
Logical fallacies are types of faulty reasoning. They can be categorized by structure (formal) or content (informal). Formal fallacies are based on a flaw in construction of a deductive argument, like claiming something is true because it might be true (appeal to probability). All represent some type of non sequitur—it just doesn’t follow. Informal fallacies are errors of reasoning, like criticizing someone’s reputation rather than his or her ideas (called an ad hominem attack).
The key formal fallacy during the pandemic seems to be the “base rate fallacy,” making a probability claim from conditional probabilities (e.g., test results when false results are a concern), without considering prior probabilities; for example, getting a positive test result from an imperfect test without considering other factors. An otherwise healthy young person without any symptoms getting a positive result is different from someone in an at-risk category with symptoms. [Before the pandemic, our granddaughter had the flu, but the test came back negative (false negative). The doctor explained she was too early to have enough viruses to come back positive.]
Multiple syllogism fallacies exist. A syllogism is a conclusion based on two premises: "all humans are bipeds, Joe is human, therefor, Joe is a biped." An often cited fallacy is: God is all powerful, god is benevolent, there is evil in this world.
On to errors of reasoning, informal fallacies. There could be a couple of hundred of these, depending on definitions. I’ll concentrate on only a few.
Equivocation is using a term with multiple meanings in a misleading way. Politicians are masters of this and design legislation to make misleading claims based on their unique definition of term, related to the budget for example: we can maintain a balanced budget with that tax cut, based on projections with unrealistic assumptions (with tables showing great results). Georgia showed innovative results by presenting a table showing a declining trend in pandemic cases and deaths just by rearranging the dates; the highest rates on the left, then progressively lower (thanks Ian Bremmer). The claim “I was being sarcastic” is another example. Trump has used this excuse multiple times, like when he claimed disinfectants could be injected to kill the virusm
An ad hominem attack is criticizing someone’s reputation rather than concentrating on the ideas presented, a favorite Trump tactic. He particularly likes to use it against journalists. If they come from the “failing New York Times,” he makes a point of calling them out—rather than answering a question. Because he likes to pontificate about the virus and his brilliant response without much knowledge, Trump seems to use this attack even more. Trump also likes to claim how incompetent Obama was and attempts to blame Obama for most problems, including pandemic difficulties like logistics.
Ad hoc attacks are rhetorical strategies such as a rationale to dismiss counter-evidence. This works well in biblical references. Pointing out archaeological evidence or logical impossibilities, the argument is: “God intervened.” There seems to be no end to ad hoc arguments. The God intervention argument seems common for demanding churches be opened immediately at 100% capacity. Referring to "Constitutional rights" to defend questionable behavioral is another, like refusing to wear a mask when required, say to enter a store.
The snow job is an obvious pandemic fallacy, the use of irrelevant facts rather than relevant evidence. This can include lies and insincere praise (aka, persuasion and deception). The value of hydroxychloroquine is praised without evidence. The snow job is universal and used widely, filling volumes of ancient to modern history, plus standard practice in a number of professions. This included the pandemic snow jobs. Trump has his list, but so do his spokespeople trying to explain why what Trump said was accurate (“alternative facts” might work). Trump also seems to expect others to offer the highest praise for his efforts—especially if they want or need anything from the federal government (perhaps a reverse snow job?).
A key informal fallacy to me is the “middle ground fallacy,” claiming each extreme side is equally right or wrong and the truth must lie in the middle—also called argument to moderation. It is important enough to write a second blog focusing on Republican politics and Trump. The “main stream” media seems to cling to this middle-ground response no matter what. This is also Median Voter Guy territory, so I try to tread carefully.
The McNamara fallacy is significant and, as an accountant, makes me cringe. McNamara as Secretary of Defense under Kennedy and Johnson used it to measure Vietnam success. Decisions were made solely on quantitative measures. He came up with “body count” to evaluate progress. With the metric, the incentives were to kill or at least “count” more dead and claiming they were Vietcong or North Vietnamese soldiers. The result was more deaths without progress. The concept also means: “if you can’t measure it, it’s not important."
In business, the top priority (maybe the only one) is profit; in the extreme, no matter what. Government and nonprofit organizations do not have (or at least shouldn’t have--matching costs with revenues can be appropriate) profit motives and rely on other measures. Governments often use techniques like program budgeting based on measures of efficiency and effectiveness. Accountants and others come up with quantitative measures (e.g., what makes a fire department effective?). This can involve big McNamara fallacies. Accountants and others try hard to come up with valid measures, but this can be difficult. For example, education success is most often measured by test results and grades. A common result is "teaching to the test" (cheating is not out of the question), not the same as promoting effective learning.
Measuring the affects of covid 19 is widespread and significant decisions are based on these measures. The accuracy of many of these are suspect and suggests a reasonable amount of caution. There is a related pandemic versus economics paradox, effective measures fighting the pandemic destroy the economy. We have measures of covid 19 cases and deaths, versus measures of unemployment rates and declining GDP. Multiple McNamara fallacies are in the works as decision-makers concentrate specifically on one or more (possibly flawed) measures.
Sunk cost fallacy is another accounting-related concept. Once the money (or other costs) has been spent, it becomes a sunk cost. If you buy tickets to an expensive performance, the money is gone whether you attend or not. Many people continue in a course of action despite evidence it’s a mistake. This includes war and pandemic response. Say a shutdown order has been in place for a month. That is a sunk cost. Say you open up after that and new covid 19 cases explode. That is a sunk cost, which could be considered a mistake. Justifying not shutting down again could be considered a sunk cost fallacy.
The privacy fallacy includes the “mind your own business” response. People’s behavior that negatively affects others and then upset by criticism give this response (which is somewhat different from the ad hoc fallacy). There is the general tension between the need for transparency and privacy, common in governments and social networks. This includes hiding or misstating pandemic information using multiple excuses.
Cherry picking means suppressing evidence or provide incomplete evidence to support your position, while ignoring evidence that does not. This is expected in legal cases with the two sides in opposition, debates, and political campaigns. Managers can justify opening up during the pandemic using this approach.
Historian's fallacy assumes that what was applicable in the past are still relevant, such as moral standards.
Post hoc ergo propter hoc ("after this, therefore because of this"), claiming that because B happened after A, B was caused by A. The rooster song happens, then the sun comes up; therefore, the rooster causes the sunrise.
Fallacy of a single cause, when in a complex environment, there are many potential causes and correlations. In a pandemic, there are complex relationships that made determining specific causes (or cures) problematic. Epidemiologists usually are fairly cautious about cause and effect, beyond what has been tested or what seems to work based on considerable experience. In centuries past, bleeding was the "cure" for multiple diseases. If the person died, he/she was beyond saving. If the person lived, it was because of the bleeding.
Red herring fallacy where evidence is misleading to make irrelevant or false inferences. Georgia presented a table of cases by day, but they were in order of large to small, not chronologically. A politician recently claimed her opponent invested in Communist China, when Chinese investments are in fact easy (perhaps expected) as part of a global investment portfolio.
Appeals to emotion are common, particularly fear during a pandemic; also appeals to flattery or ridicule. It an be easy to ridicule someone, for example, if they are photographed wearing a mask (or not). Many people claim they don't have to conform to pandemic rules because of freedom and liberty--their "constitutional rights." There are any number of other appeals: nature, novelty, poverty, tradition, or wealth.
Straw man fallacy misrepresents someone or something to insult or "disprove" their position.
Note that incentives drive results. If your told to produce a hundred widgits, expect them to be small and imperfect. If told to produce 100 pounds of widgets, they will be a limited number of large ones (also imperfect). Flawed motivations can lead to various logical fallacies.
The beauty of free markets for consumer goods: companies make money only by selling the products people want (if the only motivation is strictly profit, other problems can result). When small companies are crowded out or regulated out of existence, then we have monopoly power.
[Expect this blog to continue, either here or in other posts.]