Expertise of military leaders and national security experts


 * This essay is on Wikiversity to encourage a wide discussion of the issues it raises moderated by the Wikimedia rules that invite contributors to “be bold but not reckless,” contributing revisions written from a neutral point of view, citing credible sources -- and raising other questions and concerns on the associated '“Discuss”' page.


 * This article uses ISO 8601 dates except for References, which are controlled by standard Wikidata formatting, and "September 11, 2001". In the initial author's experience, ISO 8601 dates seem to make it easier to remember dates and to compute differences between them.

Abstract


This article claims that the goal of national security should be broadly shared peace and prosperity for the long term, and military leaders and national security experts have no intuition about that. Expert intuition is acquired by learning from frequent, rapid high-quality feedback, according to Daniel Kahneman and others. Securing anything for the long term requires learning from databases of events over a long span of history. Military leaders and national security experts are selected using other criteria, which often THREATEN broadly shared peace and prosperity in the long term.

What is expertise and how is it typically acquired?
Genuine expertise is usually acquired by learning from frequent, rapid, high quality feedback, according to a substantial body of research, including an early summary by Meehl (1954). Leaders in fields without frequent, rapid, high quality feedback are called "respect-experts" by Kahneman et al. (2021). Earlier Kahneman (2011, p. 241) wrote, "Whether professionals have a chance to develop intuitive expertise depends essentially on the quality and speed of feedback, as well as on sufficient opportunity to practice." This was the result of collaborations with Gary Klein, who had previously published work saying that experts did have genuine expertise. Their collaboration identified the essential difference between professions in which people acquire genuine expertise and those whose practitioners only have beliefs based on what they've been taught, unaltered by their experience because it lacked adequate feedback on the quality of their judgments.

"Noise audits", as described in Kahneman et al. (2021), study how variations in decisions can be attributed to factors that should and should not impact a decision. The results routinely shock experts and their managers with high levels of inconsistency, bias and susceptibility to extraneous factors like time of day, the weather, and whether their favorite sports team won the previous day. No one, especially experts and their managers, want to believe that their judgements are as inconsistent and subject to extraneous factors as they nearly always are.

For example, with asylum judges in the same jurisdiction with cases assigned at random, one judge approved 5 percent of cases while another approved 88 percent.

Similarly, Tetlock and Gardner (2015, p. 26-32) claimed that, "Until quite recently in historical terms, it was not unusual for a sick person to be better off if there was no physician available because letting an illness take its natural course was less dangerous than what a physician would inflict." That only began to change in the 1950s as Archie Cochrane, with experience as a prisoner and physician standing up to authorities in a German Prisoner of War (POW) camp, began berating his colleagues "that the only alternative to a controlled experiment that delivers real insight is an uncontrolled experiment that produces merely the illusion of insight." Most of the improvements in public health prior to the 1950s were due to improvements in public health that began in the nineteenth century. Kahneman et al. (2021) summarize research documenting examples where with some data, linear regression models with parameters estimated by least squares can outperform simple heuristics. With massive amounts of data, artificial intelligence can do even better.

How does this apply to military leaders and national security experts?

Military leaders

 * Military leaders, especially in wartime, can become experts in how to deliver death and destruction to designated targets.
 * National security experts can be skilled in convincing leading polticians and the public of the relative value of alternative courses of action to promote national security.

However, when it comes to winning wars, the record raises questions about the abilities of military leaders and national security experts.
 * Dunnigan and Martle's "General observations on two centuries of mayhem" conclude that, "Miscalculation is the most common cause of war." "Ignorance of 'the enemy' is usually a primary basis for a war." And, "[M]ost wars do not even resolve the grievances that started them."
 * When "wars do not even resolve the grievances that started them", it would seem that most military and national security authorities throughout history were respect-experts with inadequate genuine expertise for securing peace and broadly shared prosperity for the long term. In US history, this includes at least the War of 1812 (discussed below), World War I, and the US-led war in Afghanistan.
 * Even among military and national security leaders in the US, and even ignoring the US-led wars in Afghanistan, Vietnam, and Iraq, there are documented controversies about their genuine expertise. One example is the value of strategic bombing.  A Presidential Directive of 1944-09-09 led to the creation of the US Strategic Bombing Survey (USSBS). That directive and related correspondence make it clear that some leaders in the US government were wondering if strategic bombing, which had seen little use prior to WW II, was actually cost effective. Similar surveys were apparently not done of more traditional methods of projecting force, e.g., ground and naval operations, presumably because few people questioned their utility.  The USSBS found that German and Japanese war production increased throughout the war, except very near the end, after the Axis leaders knew they had lost the war. The effects of strategic bombing of Germany and Japan seem to have been similar to "The Blitz" by Germany against British cities:  Bombs typically destroyed buildings but not production equipment, which was moved to hardened sites, and the humans who survived the bombing tended to be more committed to the war effort than before. A statistical analysis of all uses of strategic bombing in history through 1999 concluded, (a) air power helped end wars when it degraded an enemy's military, but (b) targeting civilian infrastructure was not statistically significant, and (c) demanding a change in the target's leadership decreased the probability of success.

These conclusions have yet to influence public policy, but the initial author of this article is unaware of any serious research with results that contradicts these claims.

National security experts
Tetlock and Gardner (2015) described an application of the Kahneman-Klein expertise results to intelligence analysts. They described a "superforecasting" methodology, which they claimed, "performed about 30 percent better than the average for intelligence community analysts who could read intercepts and other secret data" in research funded by the Intelligence Advanced Research Projects Activity (IARPA).

This general analysis is reinforced by McChrystal's description of his experience as a brigadier general in the Pentagon leading up to the US-led invasion of Iraq in 2003. The assignment put him "in the middle of a bureaucracy that was entirely ill-equipped to deal with either the forceful secretary of defense, Donald Rumsfeld, or the highly ideological proponents of action against Iraq—most notably Rumsfeld’s deputy Paul Wolfowitz. It was often a surreal experience. ... [T]he Bush team diverged from the opinion of counterterrorism professionals and Middle East experts in its belief that Saddam Hussein was willing to tolerate the presence of al-Qaeda in Iraq. The expert assessment at the time was that Saddam Hussein and Iraq’s Ba‘ath Party were ideological opponents of al-Qaeda. We knew at the time that al-Qaeda targeted regimes like Saddam Hussein’s Ba‘ath Party. But despite experts agreeing that a Ba‘ath and al-Qaeda partnership was highly unlikely, preparations for war continued unabated through 2002 and into 2003."

Broadly shared peace and prosperity for the long term
Far more important than winning wars is securing broadly shared peace and prosperity for the long term. No one can get frequent, rapid, high-quality feedback on the long term impacts of anything, and military and national security leaders are not selected on their abilities to influence the long term.

The following subsections summarize high quality research relative to the subject of this article known to its initial author. We first consider the relative impact of violence and nonviolence on the long term outcome of conflict. We then review research on how terrorist groups end and on insurgencies.

Research on violence and nonviolence
There are substantial bodies of research on violence. Some of these use databases like the "Correlates of War" project. Others focus on nonviolence, e.g., the work of Gene Sharp and contents of the Journal of Peace Research.

The initial author of this article is unaware of much research considering violence and nonviolence together. A notable exception is Chenoweth and Stephan (2011). They created an inventory of all the major violent and nonviolent governmental change efforts of the twentieth century, over 300 of them. In brief, 25 percent of the violent revolutions in their dataset were successful, while 53 percent of the nonviolent campaigns were; see Table 1.

More important, however was their study of the change in the level of democratization: Win or lose, violent revolutions on average had little impact on the level of democratization, while nonviolence tended to improve the prospects for democracy;  see Table 2.

These are averages, and there could be exceptions. However, anyone who uses violence without more careful research than that conducted by Chenoweth and Stephan (2011) is betting against the odds and will likely lose more than they win.

How terrorist groups end


There is a massive body of publications on terrorism. However, the initial author of this article is aware of only one study that distinguished between military action and law enforcement in how terrorist groups end: Jones and Libicki (2008) compiled a database of all the terrorists groups that were active between 1968 and 2006. They found 648. Of those, 136 splintered and 244 were still active in 2006, leaving 268 that had ended. Of the ones that ended, 83 percent succumbed to rule of law, including 43 percent converting to non-violent political actors and 40 percent taken out by law enforcement. Only 20 groups, 7 percent, were defeated by military action; 10 percent won.

We should be careful not to misinterpret these results: Jones and Libicki do not say what percent of terrorist groups were confronted with military force:  If military force was used against only 7 percent of the terrorist groups, then the military would have been effective in 100 percent of the cases where it was tried. That seems unlikely. At best, we can say that 7 percent underestimates the effectiveness of the use of military force against terrorist groups, unless military force was used against all 268 terrorist groups, which also seems unlikely.

Sadly, the initial author of this work has so far not found other research that considered the relative effectiveness of law enforcement vs. military force in combatting terrorism. Hou et al. (2020) while "Introducing Extended Data on Terrorist Groups (EDTG), 1970 to 2016", included a variable indicating whether a terrorist group had been "defeated by military force or police" but did not distinguish between the two.

Insurgencies


Jones (2017) created a database of 181 insurgencies between World War II and 2015. The outcomes from the insurgencies that ended during that period are summarized in Figure 2: Roughly a third won, another third lost, and the rest ended with a negotiated settlement.

Beyond that, Jones reported that countries were more likely to be challenged by (and lose to) an insurgency where:
 * 1) most of the population is in either an in-group or an out-group, and
 * 2) there is a history of bipolar ethnic conflict or there is hyper-nationalist rhetoric from ethnic leaders in reaction to atrocities committed by one or both sides.

Effective political leaders too often exploit such grievances for short term political advantage but fail to deliver win-win reforms that reduce grievances long term.

Grievances associated with three types of conditions are particularly important: low per capita income, ethnic dominance, and religious dominance.

Countries with many ethnicities or religious groups are less likely to generate an insurgency than countries that are mostly split between two groups, one with power and the other with grievances.

Use of military force in US history
What can we gather from US history as it relates to the expertise of military leaders and national security experts? We consider here explicitly the American Revolution, the War of 1812, the Civil War, the US response to the bombings of US embassies in Kenya and Tanzania in 1998, and the current practice of killing people in foreign countries by US Remotely Piloted Vehicles (RPVs, aka "drones"), without trial by an impartial jury, as described in the Sixth Amendment to the US Constitution.

American Revolution
The American Revolution was too early to appear in the Chenoweth and Stephan database. However, their results and a careful analysis of the literature on the American Revolution suggest that most people in the US, Great Britain, and elsewhere would likely have been better off (and their descendants better off today) if the American revolutionaries had limited themselves to nonviolent action.

Many people seem to believe that the US got freedom and democracy, liberty and justice for all from the violence of the American Revolution. A careful reading of the historical documents from that period and related scholarship about those events suggests that the achievements of that revolution were more limited: The US got independence from Great Britain from the violence of the American Revolution. However, claims that the US got any improvements in democracy from that violence seems contradicted by the available evidence, as documented in the Wikiversity article on The Great American Paradox and references cited therein.

War of 1812
What can be said about the expertise of leaders who supported the US entry into the War of 1812? Then-President James Madison and former president Thomas Jefferson proclaimed that Canada could easily be incorporated into the US simply by sending the US military north to that country.

That's not what happened. Instead, the US Army in Canada killed people, destroyed property, took what they wanted at gunpoint, and 'created a Canadian national identity that had not existed before'. The Canadians returned the favor by flocking to join the British Army, which sacked Washington, DC. They also attacked Baltimore, but Fort McHenry (guarding Baltimore harbor) successfully resisted the onslaught in a battle memorialized in The Star-Spangled Banner. That's not how the War of 1812 has been described in many history books used to educate school children in the US. An honest discussion of that history might limit the options available to current US leaders and therefore would not be welcomed.

US Civil War


Figure 3 displays the dates at which different states chose sides before and during the US Civil War. Careful study of that process and other records from that period (and of the work of Chenoweth and Stephan) suggests that most people of all races and social status would likely have been better off (and their descendants would be better off today) if that war had never been fought, and the US had just let the South secede. McChrystal and Butrico (2021, p. 127) noted that "states in the Upper South, ... North Carolina, Virginia, and Tennessee, were less dependent on enslaved labor in their economies and as a result were more reticent to secede from the union. They ultimately joined the Confederacy’s rebellion only when President Lincoln ordered seventy-five thousand troops toward the South at Fort Sumter -- choosing to fight alongside their fellow southern states."

It is impossible to say with certainty what would have happened if secession had been accepted by the states that remained in the Union. The Upper South might have stayed with the Union. If war came, it would have been initiated by the Confederacy invading the Union trying to capture fugitive slaves or force the Union to continue to capture and return fugitives, per the Fugitive Slave Act of 1850. That law had been enacted as a compromise to try to keep the South from seceding earlier. It would have been more difficult to enforce after the creation of the Confederacy and might have been repealed entirely, at least as it applied to slaves escaped from the Confederacy.

In this context, it is important to note that the vast majority of the Whites in the South were poor. Many did not like having to compete with slave labor, according to Gillespie (2000). They fought for the Confederacy during the Civil War, because the North invaded their homeland. If they had been asked to invade the North, fewer would have been eager to do so. And many might have helped slaves escape.

Instead, when the Union invaded, those poor Whites were pushed to defend "their homes". The defeat of the Confederacy created animus that continues to this day.

In fact, a number of newspapers in the north were advocating that the South be allowed to secede. After the First Battle of Bull Run, some of those papers were closed by the federal government; others were destroyed by angry mobs, as described by Harris (1999).

Vietnam
In 1963 former President Eisenhower said, "I have never [communicated] with a person knowledgeable in Indochinese affairs who did not agree that had elections been held as of the time of the fighting [leading to the defeat of the French in 1954], possibly 80 per cent of the population would have voted for the Communist Ho Chi Minh". This was the universal expert consensus, at least according to the man who was President of the US 1953-1961. However, it was rarely if ever mentioned in the mainstream media of that day.

This example suggests a somewhat different message from the primary theme of this article, namely that the political environment in the US from the 1950s to the early 1970s was not conducive to an honest discussion of the situation in Vietnam. That political environment, in turn, gave US political leaders few options to do anything much different from what they actually did. For more on Vietnam, see the Wikiversity article on "Winning the War on Terror".

US response to bombing US embassies in 1998
In 1998 -- June or July -- the Taliban government of Afghanistan had reportedly agreed to extradite Osama bin Laden to Saudi Arabia to be tried for treason. The transfer was scheduled for September, because the Taliban needed time to separate bin Laden from his supporters. Then 1998-08-07 the US embassies in Kenya and Tanzania were bombed. Muslim clerics all over the world condemned bin Laden and al-Qaeda for this senseless outrage.

Then thirteen days later, 1998-08-20, the US responded by bombing a pharmaceutical plant in Sudan and al-Qaeda training camps in Afghanistan. Muslim public opinion reportedly turned 180 degrees: Bin Laden and al-Qaeda became modern Robin Hoods, standing up against what many saw as an evil empire -- the US. If everything else summarized in this article is reasonably balanced and accurate, then without those US military responses, it seems likely that bin Laden would have been extradited to Saudi Arabia in September then tried, convicted and executed there. Further, a focus on law enforcement would likely have led to the extraditions of others involved in those bombings, leading to their trials and convictions. And Islamic terrorism would likely have largely disappeared. Islamic terrorists would likely not have been able to bomb the USS Cole in 2000 nor organize the suicide mass murders of September 11, 2001. To the extent that this is accurate, it suggests that the explosive growth of Islamic terrorism following the US-led invasions of Afghanistan in 2001 and Iraq in 2003 was a result of, not in spite of, US military actions during that period.

US Drone Strikes
Has the national security of the United States been increased or threatened by the use of targeted killings using lethal unmanned aerial vehicles (UAVs or "drones")? Former intelligence analyst and drone operator Daniel Hale was sent to prison for distributing classified information documenting that nearly 90 percent of people killed by US drone strikes in a 5-month period were not the intended target.

This case raises questions about the following:
 * Has that targeted killing program made the US safer? Or has it manufactured more enemies for us than it eliminated? The supporters of the program would presumably insist that it has made the US safer. However, the evidence summarized elsewhere in this article raises questions about that judgment. On a different but related issue, Stanley McChrystal, commanding general of US-led forces in Afghanistan, 2009-2010, said that "our response to mitigating risk [with 'night raids' in Afghanistan] instead augmented the dangers we wanted to avoid."
 * What portion of information classified by US government officials is classified to keep it from the US public -- or even the US Congress? Per the US Supreme Court decision in US v. Reynolds (1952), no federal judge in the US can question a claim of national security by a federal government official. That exemption means that massive violations of US and international law can be routinely kept from judicial scrutiny, and anyone who exposes such crimes can expect to spend time in prison, like Daniel Hale.  Connelly (2023) suggests that government secrecy may actually threaten US national security, because it has allowed government officials to clandestinely provoke foreign entities into attacks that were denounced as "unprovoked". The public and Congress were then stampeded into authorizing questionable responses without adequate deliberation. Connelly also claims, "There is nothing more dangerous—both to itself, and to others—than a nuclear-armed superpower that is not even answerable to its own people." He further says, "We should actually be more alarmed by the very real risk that all our preparations for war will result in the United States’ provoking just such an attack, or accidentally launching one of our own." For previous discussions of US government secrecy, see Moynihan et al. (1997) and Graves (2014).

No conclusions nor implications here
The conclusions and implications of these observations are vast and highly controversial. We ask only one question here, combining the observations of Connelly (2023) and Tetlock and Gardner (2015):


 * How might US law be changed to increase national security to benefit the bottom 99% of the US and global populations, perhaps at the expense of people currently exercising power?

Further analyses will be deferred to other articles in hopes of separating issues and focusing attention in this article on questions raised by research on expertise as it applies to military leaders and national security experts.