In the wake of the recent Friends of Syria conference, the United States and Middle Eastern powers that include Turkey, Qatar and Saudi Arabia are stepping up aid to armed resistance groups in Syria. Under American leadership, the conference pledged $100 million to provide salary payments to rebel fighters.

Whatever the humanitarian intentions, this strategy, along with discussions of “safe zones” and “non-lethal aid,” is misguided at best, and counterproductive at worst. For all the talk about safeguarding civilians, the proposals are far more likely to escalate violence than to reduce civilian casualties.

To appreciate why, look at events in Syria: An authoritarian regime is engaged in brutal repression and large-scale human rights violations. According to the United Nations, the death toll exceeds 9,000, with thousands more wounded.

But arguments for direct assistance to the rebels miss the political complexities within Syria. On one side, large sections of the country oppose President Bashar al-Asad, and a significant proportion back efforts of armed rebel groups to topple his government. In cities like Homs and Dir‘a, the Syrian army’s brutality has consolidated support for rebels.

Elsewhere, however, in major cities like Damascus and Aleppo, the picture is more mixed. Some neighborhoods support a network of anti-government protest organizers known as the Local Coordination Committees. But large sectors of society -- including most minority groups like the Christians, Druze and ‘Alawis, along with much of the middle class and business owners -- have, for the most part, withheld support from the rebels. The most likely explanation is that they fear more instability and violence, as well as some possible outcomes, like rule by the Muslim Brothers.

These internal divisions belie the argument that all segments of the civilian population would welcome intervention.

Given these realities, international actions like those discussed at the Friends of Syria meeting risk worsening the situation. Aid to opposition forces is fungible; even when it is “non-lethal” or financial, it amounts to arming the rebels and taking sides in a civil war. Those advocating such measures include Sunni-majority countries whose assistance against the ‘Alawi-dominated regime could fuel sectarian and ethnic tensions.

And establishing “humanitarian corridors” on Syrian soil would require defending them to protect fleeing civilians. If foreign forces offered these defenses and came under threat, escalation into direct intervention would be likely. Further, such internationally defended lines would be tempting grounds for armed opponents to retreat behind after their attacks -- another scenario that would increase rather than reduce the risk to civilians.

The prospect of intervention must seem welcome to protesters subjected to brutal repression by the government. But the principal requirement of an intervention on humanitarian grounds is the prospect, on balance, that it will offer greater protection to vulnerable civilian populations. And this is absent in Syria.

So what else is at stake in the calls for intervention? If the involvement of outside actors turns a civil conflict in Syria into a proxy war against Syria’s government and its allies -- namely Iran and Russia -- then intervention would bolster pro-Western powers at the expense of Syrian civilians. And such action would continue to erode international constraints on the use of force, especially those based on national sovereignty.

Ultimately, the best way to reduce violence is to pursue negotiations for a political transition that would include rather than explicitly threaten the Asad government. Given the mortal fears of communities on each side of the conflict, the first goal has to be making clear that all groups have a future in a new Syria.

The six-point plan offered by Kofi Annan, the United Nations intermediary, is a good starting point. But both sides have to treat a ceasefire seriously, and any arms embargo would have to apply equally to each party. Crucially, real negotiations would have to include Iran and Russia. Both have stakes in the Asad government; their involvement in an inclusive mediation process could set the stage for concessions by the government.

Some will argue that we shouldn’t engage with the Syrian government or its backers. But further isolation tells the Asad government and its social constituencies that their only options are victory through mass violence or annihilation.

By relying exclusively on coercion through sanctions and threats, the practical effect of the current American approach has been to squeeze out all other diplomatic options and to make a proxy war (with local and international players on both sides) the only remaining possibility.

If we are really interested in protecting the civilian population -- rather than using this as a strategic opportunity to flip regional alliances -- the benefits of a negotiated transition are clear. It may not reinforce our geopolitical position, but it will help safeguard ordinary Syrians caught in the crossfire.