The United States has a problem unique to history's greatest powers. Our wars are often too cheap. Air campaigns like the 2011 campaign in Libya, the ongoing one against the Islamic State in Syria and Iraq, and the drone strikes that periodically target militants in Pakistan, Yemen, and Somalia sacrifice few if any American lives and spend the federal equivalent of loose change.

Because these wars risk so little, or seem to, we have grown too fond of them. We make war without much regard for whether it is worthwhile. Recent U.S. decisions to bomb countries bear less resemblance to the struggle between branches of government that the constitution anticipates than to one-click shopping online, where low upfront cost and ease of delivery encourage whimsical choices uninhibited by debate about value.

Drones, or unmanned aerial vehicles, as their makers prefer, did not cause this circumstance. It results from the United States' good fortune: wealth, geography that keeps enemies distant, military superiority, and technological prowess. These advantages have long tempted Americans to use military technology, especially airstrikes, as a quick fix for distant political challenges. In recent decades, progress in surveillance and targeting capability and the weakness of U.S. rivals have enhanced the temptation. Drones, by making war seem even cheaper, just exacerbate the problem. In that sense, they are a quintessentially U.S. weapon.

Ours is a good problem to have. The ability to fight war without much risk is a privilege of fortune. Other states would eagerly suffer that malady in the same way that most poor people would brave the temptation to overspend on shopping by becoming fabulously wealthy.

The trouble is that airstrikes and other quick applications of military force are rarely as cheap as they first appear. They tend to cause unanticipated trouble and begin conflicts without winning them. Escalation to more costly warfare then beckons. Drone strikes may prove to be especially misleading this way. Their benefits come fast and are straightforward. Most strikes bring reports of dead terrorists or insurgents, and their disrupted plans are easily imagined. The costs—especially blowback measured in violent anti-American sentiment and pressure toward escalation — arrive gradually and less discernibly.

(U.S. Air Force photo/Senior Airman Nadine Y. Barclay)

The escalation danger is especially underappreciated. As has often been the case with strategic airpower, drone strikes tend to achieve some tactical success without achieving war's political ends. The choice might then be to give up or double down. An initially low price-tag might turn out to be a down-payment.

These dangers recommend a more cautious approach to drone warfare. We should avoid limited military strikes in circumstances when we are unwilling to risk more. We should resurrect the notion that wars, even those that seem cheap and quick, should be hard to start.

This essay explores drones' unappreciated costs and explains how they skew collective decision-making about war. But it should not be confused for an anti-drone essay. The complaint here that we launch drone strikes too easily does not imply that the right amount of strikes is none. Besides, efforts to ban military technology, like stopping drone production or outlawing armed drones, generally fail. Technical adjustments and slippery definitions allow weapons to evade bans. It is more feasible to restrict Presidents' standing authority to launch strikes.

Drones also serve a variety of useful military functions besides the controversial hunter-killer role. Predators, Reapers and other armed drones get more attention, but small surveillance scouts amount to the vast majority of U.S. military drones. Surveillance will likely remain drones' primary military function, even if, as some suggest, drones replace U.S. long-range bombers or mid-air refueling aircraft. Drones' combination of range, relatively low cost (largely because they are unmanned), and loitering capability make them well-suited for surveillance, which is now a component of almost every modern U.S. military mission.

U.S. drone strikes occur mostly amid conventional wars. According to the Bureau of Investigative Journalism, the United States conducted 145 drone strikes in Libya during the 2011 war there and nearly 1,100 drone strikes in Iraq Afghanistan from 2008 to 2012 (figures for prior years are not publicly available). The vast majority of those came in Afghanistan, probably due to its remote terrain. Today U.S. drone strikes occur regularly as part of the campaign against ISIS in Iraq and Syria, though drones there mostly serve to find targets for manned aircraft to attack.

According to the New America Foundation, the United States has launched 399 drone strikes in Pakistan and 128 in Yemen, with most of the strikes coming in the last five years and slowing over the last several. Another 9 to 13 U.S. strikes targeted people in Somalia, by the Bureau's count. In 2006, one U.S. drone strike occurred in the Philippines. Another recently killed a terrorist in Libya. More are likely there given reports that U.S. officials are negotiating for basing rights in nearby states.

Before getting further into these cases, it's worth considering the martial context in which they occur. The United States is conducting airstrikes in seven countries, using regular ground forces in one or two of them, depending on how you count, and employing special operations forces for raids, targeting or training in many more. We are working, in Syria, on overthrowing our third middle-eastern government in the last decade and half.

Even leaving aside naval patrols and U.S. garrisons in allied states, that is a lot of military activity. But it is not particularly controversial or important in current U.S. politics, where, as is generally the case, pocketbook issues dominate. Despite all this war, President Obama is criticized in national security debates mostly for being too dovish. His successor, whether it is Hillary Clinton or any Republican besides Rand Paul, will likely be more hawkish. Widespread, relatively low-level warfare that limits risk to U.S. personnel has become normal and what Washington's moderates support.

How did that occur? For starters, keep in mind that warfare does not escape the effect of price. If a war is likely to be disastrously expensive, in the human or financial sense, most states avoid it. That is how deterrence works. Make war cheap and easy and arguments for it flower. That is why powerful states generally make war on weak rivals, like Libya, Iraq, and Georgia, not each other, and why drones make war more likely.

Historically, three factors have encouraged Americans to seek technological shortcuts to winning wars. First, as historian Alex Roland observes, labor has tended be relatively scarce in the United States and capital relatively abundant. The U.S. military, as a result, tends to substitute capital for labor, building technology to deliver firepower and limit risk to men.

Second, because United States has for more than a century faced little threat of invasion or civil war, it fights only distant wars where ships, aircraft, and missiles that can strike at distance are especially valuable. Moreover, because the nation's physical safety is not directly at risk in these wars, it is even harder to justify sacrificing lives for them.

Third, Americans' liberalism encourages U.S. leaders to justify wars as means to advance liberal values. That typically involves casting foreign leaders as oppressors of otherwise democratic people. Those rationales militate against indiscriminate violence and invite a search for technologies that attack foreign militaries but spare civilians.

These factors produce what Eliot Cohen calls the mystique of American airpower. This is the hope, linked to the birth of the independent U.S. Air Force, though few historical successes, that airstrikes alone will win wars without much risk to our troops or foreign civilians. Airpower's mystique seduces by claiming to reconcile war and liberal values, denying General William Tecumseh Sherman's dictum that war is cruelty.

In the 1990s, the marketing of the Gulf War victory renewed infatuation with strategic airpower. The Clinton administration's bombing campaigns in the Balkans and Iraq labored to avoid casualties but promised big results. Lately, special operations forces, or at least their boosters, offer a similar story. The suggestion is that these forces are not just highly-trained units, especially good at working with other militaries or raids, but a transformative instrument that can organize foreign states by training their militaries and destroy insurgencies by killing or capturing their leaders.

Several recent developments encourage this old American infatuation with technology-centric warfare. First, a surveillance and precision targeting revolution, realized over decades, has made aircraft exponentially more effective at destroying fixed targets. Computing and communication gains enable surveillance systems on various platforms to find and precisely locate targets and instantly transmit that information so that bombs or missiles can use laser guidance or the global positioning system to rapidly destroy them. Aircraft can do all this themselves. Drones add the ability to sit and watch for extensive periods. Accuracy compensates their small payload—few shots are needed if misses are rare.

Second, the Obama administration inherited and enhanced a supercharged imperial presidency. Congress is full of members who have served entirely since 2001 and are thus unaccustomed to using their constitutionally-granted war powers. They allow this administration, like the last one, to use the 2001 Authorization of Military Force law to launch airstrikes more or less wherever they want, or as was the case with Libya, to evade authorization by simply pretending the war is not a war but a "kinetic military action."

Third, jihadists, prone to proclaim allegiance to al Qaeda and lately the Islamic State, remain a factor in several Islamist insurgencies. Some focus on local enemies and make no obvious effort to target Americans. But their stated allegiance to those who do makes not bombing them a tough sell for elected U.S. officials.

The deficit and war weariness lately curtail those pushes toward intervention. Though war funds, unlike regular military spending, are not subject to the spending caps put in place in 2011, deficit concerns restrain them somewhat. More important, the tragic results of the occupations of Iraq and Afghanistan created historically high public opposition to ground wars. Even Senators busy assuring us that the Islamic State will be on our shores in no time deny that U.S. troops should go fight them in Syria and Iraq.

So on top of a proclivity to try to win wars with more technology and less manpower, we lately gained heightened capacity to use force discretely (discrete is a relative term here), and a Congress happy to let presidents do so without their permission. On the other hand, debt and increased anti-ground war sentiment limit the nation's willingness to pay for war.

The result of these conflicting forces is limited liability interventions and U.S. military actions woefully incommensurate with their stated goals, which retain their typically excessive ambitions. To highlight only the latest examples, the Obama administration now argues that bombing the Islamic State and training and arming a few rebels in Syria will empower them to win and rule decently there, despite the dearth of trainable, decent rebels. The President's big speech on terrorism at West Point in spring 2014 claimed that a bit more military-to-military training would shore weak states against collapse and thus terrorism, despite the absence of credible arguments that state failure results from insufficient military training or reliably causes terrorism.

Drones make war seem even more remote and cheap. They heighten this tendency to see war as a problem awaiting a technical solution. So what, you might ask. Even if drone strikes and other limited military actions do not accomplish as much as advertised, they might do enough good to justify their limited cost. Why not try a few strikes, if, as is the case, the cost is sustainably low?

It's true that drones provide some counterterrorism benefit, even though it's hard to measure. According to New America, U.S. drone strikes in Yemen, Pakistan and Somalia have killed somewhere between 3,000 and 4,800 people, of which about 400 were civilians. Even if the majority of the dead combatants were insurgents focused on local targets, not terrorists plotting against westerners, the strikes likely disrupted their operations, if only by making it hard for them to move, assemble and cooperate. Some captured al Qaeda communications suggest as much. Armed drones have gone a long way toward denying terrorists haven in remote places, even though the political will to attack them there, one way or another, probably deserves greater credit.

Still, the fact that a policy is affordable and has some benefit does not establish its wisdom. You still need to weigh its costs. Lots of foolish acts are sustainable, at least for a while, especially when you're rich. That does not make them a good idea. Unsustainable policies, which tend to produce politics conducive to their end, may ultimately be less problematic than sustainably dumb ones, which bleed you slowly, but not enough to generate opposition or debate.

Drone strikes are especially likely to fit in that slow bleed category because their costs are low, compared to most wars, but not as low as they seem. They initially either occur downrange, in the form of dead people whose families can't vote, or in the future, as abstractions like resentment. Because these costs are slow to arrive and obscure, while the benefits are relatively concrete and immediate, drone strikes have a specious attraction. That makes them especially resistant to judicious debate.

Before discussing drones' underappreciated costs, it is worth noting the absence here of one charge often leveled against drones. That is the idea that they are especially prone to killing noncombatants. As the Drone Task Force Report released last year by the Stimson Center points out, drones are relatively precise compared to other long-range weapons. Their danger to civilians comes less from imprecision than by serving the myth that you have war without killing the innocent. Of course the Central Intelligence Agency and Joint Special Operations Command, which operates armed drones for the U.S. military, should minimize the number of innocent civilians that drone strikes kill. But the best way to protect civilians is to have fewer wars.

Even drone strikes' direct costs are somewhat disguised. The Stimson Report notes that U.S. drones' financial advantage vis-à-vis manned aircraft probably declines over their operational lives. Drones' procurement costs are generally lower than those of manned fighters or bombers. But drones are often pricier to operate, largely because they require more human support on the ground.

A larger delayed cost of drones is blowback. The standard form occurs when those angered by strikes support or even become terrorists. If drone strikes reliably produce that result, they are likely to generate an endless supply of terrorists to hunt. Blowback might also mean support for anti-US insurgents, like al-Shabaab or the Taliban, or simply diplomatic consequences, such as disinclination in foreign capitals to cooperate with U.S. requests for intelligence.

Blowback might arrive as a self-fulfilling prophecy. Fear that Islamist insurgents will sooner or later become terrorists that attack Americans encourages drone strikes that turn them into what we feared. It's likely that has occurred to some extent in Yemen, Somalia and Pakistan, where strikes have targeted an amorphous mix of insurgents and terrorists in their midst.

Most examinations of drone strikes discuss blowback. But the difficulty of observing and measuring blowback makes it likely to be undervalued in whatever cost-benefit analysis guides government thinking about strikes. One indication of blowback is that terrorists arrested in the United States frequently cite anger about U.S. wars and drone strikes as their motive. Another sort of evidence for it comes from studies of past bombing campaigns, like the post-World War II Strategic Bombing Survey, that show how bombing people tends increase their sense of nationalistic or ethnic solidarity and heighten their support for resistance against the attacker.

A third underappreciated danger of drone strikes is the risk of wider war. The strikes' seductively low cost might pull us into wars and then, by failing to win them, invite escalation. Drone strike advocates typically dismiss this concern, contending that the alternative is ground forces or some other, more costly action. But the alternative is often peace. Without armed drones, the United States would likely have conducted special operations raids or conventional airstrikes in Yemen, Somali, or Pakistan. But given the greater risks that those options entail and the host nation's more limited ability to deny allowing them, the number of strikes would likely have been far lower, and they might by now have ended.

Drone strikes risk escalation by failing to solve the political conflicts that produce the terrorists that they target. Even if they successfully disrupt terrorist operations, they will rarely seem to succeed. That perceived failure, along with drone advocates' tendency to overpromise, might create domestic pressure to continue attacks indefinitely, increase the volume of strikes or to escalate to wider war.

True, none of the U.S. drone campaigns have yet seen escalation to conventional airstrikes or ground war. But it is possible that the drone strikes in Pakistan and Yemen increased in volume in prior years partly due to their failures. Note that we haven't ceased drone attacks in any nation, besides after that single attack in the Philippines. The next President, presented with drones' failure in Yemen, Libya or wherever, may be convinced to widen the war.

The final risk of drone warfare is to further erode the checks and balances conducive to sensible foreign policy. As noted, Congress was busy abdicating its war powers well before armed drones were available. But drones seemingly low cost make Congress even less interested in policing presidential action.

Where wars impose few evident costs on the public, it lacks incentive to pay attention and form strong opinions. People do not press their elected leaders much, leaving politicians with little reason to protect their constituents by overseeing war and checking presidential power. The Vietnam and Iraq Wars are exceptions where the human and financial costs grew to a point where they became quite important electorally. Congressional oversight, as a result, increased over the course of those wars.

The Libya War is at the other end of the spectrum. Unless we count the Benghazi attack that occurred amid the chaos that Qaddafi's fall produced, the war killed no U.S. personnel. The United States spent about two billion dollars on the bombing. That amounted to less than half of one percent of that year's U.S. military spending, which itself was then between four and five percent of gross domestic product. The Obama administration funded the effort through transferring existing Pentagon funds.

Because the direct consequences of war were negligible, there was no Congressional appropriation, let alone an authorization vote. Congress barely debated the war. Its rationales suffered little scrutiny. For example, the administration made dubious claims about the vast humanitarian value of aiding rebels and neglected the observation, made by a few academic critics, that Qaddafi's fall was likely to bring long-term chaos destructive to humanitarian ends, as is now occurring.

Drone wars are similar. The administration barely offers public arguments for strikes in Yemen and Somalia, let alone particular strikes. Secrecy and a lack of obvious cost limits demand for justifications. To the extent they exist, they get almost no Congressional scrutiny, at least in open hearings. For the public, having an informed opinion about the merits of strikes becomes nearly impossible.

Edward Corwin long ago described the Constitution's division of powers as an invitation to struggle for the privilege of directing U.S. foreign policy. But where costs distribute such that no group suffers too much, Congress does not accept the invitation. Checks and balances fail. We get a kind of autocratic foreign policy, where public disinterest allows presidents to act unilaterally. Because these decisions avoid the examination and debate produced by competing interests, they are more likely to be bad ones.

Consider that even the Republican hostility to the Obama presidency has not caused them to use their majority in both Houses to restrict his war powers. Senator Marco Rubio, among others, even criticizes the President for not claiming even broader powers to ignore Congress. Such habits of undue deference might be difficult to reverse even if the present wars end.

Besides the deferential Congress, we now have a national security establishment that has largely come to believe that the exigencies of our foreign policy require that the president be able to start wars as quickly as possible. Everyone, especially in the Pentagon, behaves as if Presidents can never have too many martial options, even though, as Bernard Brodie once noted, the basic logic of our constitutional system says otherwise. In Washington, it sounds quaint to say that our government—especially when it comes to making war in distant places — is designed more to provoke conflict conducive to caution than the efficient dispatch of bombs and soldiers.

Happily, the United States seems likely to remain safe, rich, technologically proficient and militarily superior to rivals for the time being. As a result, we are likely to keep casually launching limited strikes in pursuit of big objectives. But we can mitigate the problem somewhat by encouraging Congress to restrict the president's war powers and by trying to realize the full cost of wars up front.

A simple step to establish some Congressional control over drone strikes is ending the 2001 Authorization of Military Force (AUMF). That is the strikes' primary legal basis. Contrary to Congress' current inclination, which is to update the AUMF with broad new authority more tailored to the times, we should simply end it or replace it with strict temporal and geographic limits on where strikes can occur. If a President wants to attack a new group or an old group in a new country, he or she should get a new authorization from Congress. That might force a bit of informative debate.

A related reform is to stop funding wars through overseas contingency operation (OCO) appropriations. These have become standing slush funds that allow the administration to start various little wars, including new ones, without much guff from Congress. The AUMF is like standing permission for presidents to have these wars. OCO is like a credit card that funds them. Likewise, although it will do little to inhibit drone strikes, we could use a law that says our ongoing wars must be funded in the coming year by a tax increase or offsetting spending. That would force some slight tradeoffs and spark debate about priorities.

Congress might also reform the convoluted law governing covert wars. But an even better reform would be to stop having them. Our drone strikes are only secret in the legal sense that hinders oversight.

Probably the most powerful way to restrain feckless wars is to restore old-fashioned mistrust of promises that technology, genius generals, or any other invention can deliver military victories without sacrifice. We should try to keep in mind that bombing and killing people, even precisely, tends to produce unanticipated trouble. Those unwilling to pay much for wars should probably avoid them.

Benjamin H. Friedman is a research fellow in defense and homeland security studies at the Cato Institute and an adjunct lecturer at George Washington's Elliott School of International Affairs.