The crescendo from America’s widespread Black Lives Matter protests has reached the halls of the Pentagon. In short order, the Air Force and Navy held a series of searching town halls about racism within the ranks; Defense Secretary Mark T. Esper banned the display of Confederate flags at military installations and created a defense board to tackle questions of diversity and inclusion; and the new $741 billion defense authorization bill adopted race-blind promotion policies and the renaming of 10 bases named for Confederate figures. (President Trump has threatened to veto the bill.)

These steps toward greater inclusion will pay dividends on future battlefields. As I show in my new book, “Divided Armies,” victory on the battlefield over the past 200 years has usually gone to the most inclusive armies, not the largest or best-equipped ones. Inclusion, in other words, is good for military effectiveness.

Here’s how I studied military inclusivity around the world

Building inclusive armies isn’t simply a moral question. My research shows that inclusive armies fight harder, suffer lower rates of desertion and defection, and exhibit more creative problem-solving on complex battlefields than armies drawn from marginalized or repressed groups.

To see why inclusion provides a war-winning edge, it can be useful to set the U.S. military in a comparative context. I gathered data on the ethnic and racial inequalities within nearly 850 armies in 250 wars fought since 1800. I determined ethnic inequality by measuring two aspects of each army: the demographic share represented by each ethnic group in the army, and the way in which the government treated each ethnic group in the prewar era. For this measure, I recorded whether the country was inclusive toward the ethnic group, discriminatory or repressive.

Combining the two aspects creates an index that runs from 0 (perfect equality) to 1 (perfect inequality). Imagine, for example, an army with three ethnic groups divided between a favored group (50 percent of soldiers), a marginalized group (25 percent) and a repressed group (25 percent). This army would receive a military inequality score of 0.38.

This means it is not enough for an army to be diverse — a country must also extend equal and full citizenship to the constituent ethnic groups in the army to reap the benefits of inclusion. The closer an army inches toward a score of 1, the worse its expected battlefield performance, as rising inequality saps the military cohesion necessary to wage modern war effectively.

How inequality makes armies less effective

Why is inequality in the military so harmful? Inequality erodes battlefield performance in four specific ways.

First, shutting certain groups out of senior decision-making robs the army of diverse thinking, reinforcing a “group think” mentality that tends to lead to deadly conservatism on the battlefield.

By excluding certain voices, militaries saddled with high inequality become locked into predictable, usually second-best, tactics and approaches that leave them vulnerable to enemies.

Second, inequality means different ethnic groups have varying status, and this sows tensions and erodes trust among soldiers. In turn, this can hobble cooperation and coordination on chaotic battlefields.

Third, soldiers drawn from targeted ethnic groups may suffer a loss of morale, making them less likely to fight for a government they distrust — and more likely to surrender.

Fourth, prior exposure to government-directed discrimination or violence builds strong bonds among soldiers within the same ethnic group. Those who feel discriminated against may use these ties to organize escape or otherwise subvert military authorities. Military commanders may then be forced to devote greater resources to policing their own soldiers, creating new vulnerabilities that enemies could potentially exploit.

The historical record shows that diverse armies fare better

The perils of inequality are borne out by historical evidence. Since 1800, armies with high rates of inequality have done poorly, according to all kinds of measures of battlefield performance.

In my statistical analysis, moving from low to high inequality, for example, increases the odds by 50 percent that an army will suffer more casualties than it inflicts on the other side. Similarly, the likelihood that large numbers of soldiers will desert the battlefield increases by a whopping 70 percent when looking at low-inequality vs. high-inequality armies. Defection — when soldiers switch sides and take up arms against their former comrades — is 60 percent more likely at high inequality than low.

And armies with medium or greater inequality are far more likely to turn their weapons against their own soldiers. During World War II, for instance, the Soviet Union’s Red Army, one of the world’s most polyglot militaries, resorted to killing an estimated 158,000 of its own soldiers to force them to stand firm against Nazi forces.

We sometimes catch glimpses of armies rebuilding themselves during wartime. During the Korean War, for example, the desegregation of the U.S. Army improved the casualty survival rates of integrated regiments over whites-only units fielded early in the war.

Progress on equality can be reversed, however, with sharp consequences for battlefield conduct. Here’s an example. The ragtag Mahdist state’s inclusive armies managed to defeat superior Anglo-Egyptian forces in 1881-1885. But the death of the Mahdi leader in 1885 led to a recasting of the state and its military around a narrow ethnocratic vision that violently subjugated most of Sudan’s population. The result? The bitterly divided Mahdist forces, often fighting under duress, were annihilated at Omdurman in 1898 — one of the most lopsided defeats in history.

Why this matters for the U.S. military

In 2020, as the United States gears up for a return to “great-power competition,” the margin for error against a newly assertive China and revisionist Russia will continue to shrink. Building, retaining and fielding a diverse and truly inclusive military will remain a central challenge for the Defense Department.

Removing the structural barriers to inclusion will be one important area of focus. Nonwhite personnel continue to be underrepresented at senior levels, a problem that worsens as one advances up the ranks. Women and nonwhite personnel also experience slower rates of promotion. Progress toward inclusion has also proceeded in fits and starts. Since 1945, the U.S. military has launched at least a half-dozen initiatives to stamp out racial inequalities.

For the U.S. military, investing in its people, not just military hardware, and redoubling a commitment to meaningful inclusion would bolster the war-fighting advantages of diversity.

Jason Lyall (@jaylyall_red5) is the James Wright chair of transnational studies and associate professor of government at Dartmouth College. He is the author of “Divided Armies: Inequality and Battlefield Performance in Modern War” (Princeton University Press, 2020).