Subsidizing Peace

The following is a guest post from SUNY-Buffalo political scientist Phil Arena.


Observers of international politics can be forgiven for being skeptical about ceasefire between Israel and Hamas recently brokered by the United States and Egypt.  In substantive terms, it is nearly identical to the ceasefire reached at the end of Operation Cast Lead in 2008-2009.  Moreover, as Joshua Tucker discussed just the other day, recent work on conflict management paints a pretty grim picture.  Though mediators might be able to achieve genuine conflict resolution under some conditions (see  here), those conditions appear to be quite rare (see here and here).  The conditions under which conflict management can be achieved are more easily met, but simple management of conflict appears to merely delay the inevitable (see here, here, and here).

However, I think there is a case for optimism.  (Those who know me well will understand how unusual it is for me to write those words.)

I say this for two reasons.  First, though I largely agree with the conclusions of extant work on conflict management, I think there theoretical reasons to believe that these unhappy conclusions need not apply under certain conditions.  Second, I think those conditions may well apply in the current case.

Anna Pechenkina and I have a working paper in which we demonstrate formally that third parties can bring lasting peace by providing conditional subsidies, even if they fail to resolve the underlying information or commitment problems that can cause conflict.  They do so, quite simply, by raising the opportunity cost of war.  If you promise to throw a ton of money at people, but only if they don’t shoot each other, there’s a pretty good chance they won’t shoot each other.

That is, we think, exactly what the US did with Israel and Egypt.  In the second half of our paper, we argue that the reason why Israel and Egypt have seen nearly 40 years of peace, after fighting five interstate wars between 1948 and 1973, has less to do with Camp David and more to do with the dramatic increase in foreign aid outlays that began almost immediately after the Yom Kippur War.

Analyzing all contiguous dyads involving Israel from 1948 to 2001, we demonstrate that the number of Militarized Interstate Disputes (MIDs) involving the use of force that are initiated in any given dyad in any given year is positively associated with parity (which exacerbates information problems) and anticipated future shifts in power (which can induce commitment problems), but these relationships are conditioned by the total amount of economic aid provided by the United States.  In the presence of sufficiently high levels of foreign aid, the effects of parity and shifts in power disappear, and the predicted number of MIDs drops precipitously.

Of course, correlation is not causation.  To strengthen the case that the reason why Israel and Egypt have not fought another war since 1973 is because of the foreign aid the US has provided, and that the US has provided aid for precisely this reason (at least with respect to Egypt – there are of course other reasons for the US to provide aid to Israel), we also analyze the total amount of aid provided by the US to any given dyad in any given year.  We find that the amount of money provided by the US is significantly greater in years where observable factors would suggest that the risk of conflict would otherwise be great – but only after 1973, when we argue that the US experienced a large exogenous shock to its interest in stability (in the form of the OPEC oil embargo placed enacted to punish the US for its support of Israel in the Yom Kippur War.)

We also note that recent events support this interpretation.  Earlier this year, after Egypt’s ruling Supreme Council of the Armed Forces (SCAF) began cracking down on pro-democracy non-governmental organizations in Cairo, Representative Ron Paul and Senator Rand Paul attempted (unsuccessfully) to cut off all aid to Egypt (see here). The Muslim Brotherhood’s political arm immediately threatened to “revisit” the 1979 treaty with Israel if aid was to be cut off (see here), explicitly linking peace to the provision of foreign aid.  As the New York Times says, “Egyptians have long considered American aid as a kind of payment for preserving the peace despite the popular resentment of Israel.”

It is therefore quite interesting that though the recent ceasefire between Israel and Hamas is very similar to the one that followed Operation Cast Lead, this time around, as Haaretz reports, “Netanyahu received American compensation for his agreement to a cease-fire. President Obama…promised to increase U.S. military assistance to Israel, especially regarding the prevention of arms smuggling into Gaza…and to purchase more Iron Dome and other anti-missile systems.  According to the understandings, Israel has undertaken not to launch any attacks on Gaza – by land, sea or air – and to stop the assassinations of the heads of Palestinian militant groups and not invade any Palestinian-held land.”

To sum up, we see little reason to believe that the underlying issues have been resolved.  But conflict management can bring sustained peace even when it fails to achieve that lofty goal.  Simply by manipulating actors’ material incentives, specifically by raising the opportunity costs of war, third parties can bring peace in the long term.  We think the remarkable transformation of Israel-Egyptian relations serves as an example of this strategy.  That the US seems to now be mimicking the approach that brought peace between Israel and Egypt suggests that, this time, the ceasefire may bring more than a temporary reprieve.

Of course, that assumes not only that the United States will continue to offer even larger foreign aid disbursements to Israel than it has provided in the past, but also that it will make clear that (the increase in) foreign aid will go away if Israel renews hostilities in Gaza, and that Israel will consider such threats credible.  Threats to withdraw or reduce aid disbursements to Israel should Israel target Hamas again may or may not be as credible as threats to withdraw aid to Egypt should Egypt attack Israel.  If my (uncharacteristic) optimism proves unwarranted here, though, I would not interpret that as further evidence that mediation merely delays the inevitable.  The culprit will not be a fundamental shortcoming of conflict management so much as the inability of the US to credibly commit in this particular instance to subsidizing peace and punishing the breakdown thereof.

24 Responses to Subsidizing Peace

  1. Vladimir November 23, 2012 at 11:47 am #

    There is also an argument which has been applied to the US-Egypt case that the subjects of mediation attempt to exploit the mediator to maximize their gains from peace.

    J.J. Gillespie and Max H. Bazerman, “Parasitic Integration: Win-Win Agreements Containing Losers,” Negotiation Journal 13, 3 (1997: 271-283

    • Phil November 23, 2012 at 4:08 pm #

      Thanks for the comment, Vladimir. I have seen similar arguments made, especially about peacekeeping,b ut I haven’t seen that particular paper. I will check it out.

  2. Felix November 23, 2012 at 12:26 pm #

    That’s an interesting an argument, and I’m definitely going to check out that working paper. Two immediate reactions, though:

    First, if I understand you correctly, you’re defining “costs” in material terms (foreign aid; military and human costs of warfare). Would it change the model if you included non-material costs such as reputation, perception of security, and or electoral costs? (Probably: yes, but that would make the model a lot more complex, right?)

    Second, I don’t find Hamas anywhere in your application of the model to the current Israel-Hamas conflict. Wouldn’t you need to provide subsidies for both sides to change their respective utility expectation of war? If you only provide Israel with “tons of aid money” but not Hamas, they don’t have any externally induced incentive not to go to war again. They might even be more inclined to start bombing again, since they could perceive the settlement (ceasefire+aid money for Israel but nor for them) as injust.

    Then, in turn, Israel wouldn’t respond at first, because their costs are still somewhat higher now due to the US aid money. But eventually, reputation, security and electoral considerations (that’s why I asked for including these calculations in my first point) might cause them to respond anyway, because they outweigh the benefits from US aid.

    • Phil November 23, 2012 at 1:36 pm #

      Great questions, Felix.

      I wasn’t clear about this in the post, but the model treats the costs of war simply as the subjective loss of utility associated with fighting. That can reflect lots of things besides material costs. The *increase* in the cost of war that the third party brings about is assumed to be strictly due to material inducements, but in principle, the logic would not change if the third party could increase the expected loss of utility associated with fighting through some other mechanism.

      You are right that I didn’t say much about Hamas. According to the model, it doesn’t actually matter which side is given the subsidies. The intuition is that if Israel knows that Hamas has every bit as much of an incentive to fire rockets now as it did before (perhaps even more so, due to resentment of the terms of the ceasefire), then Israel must choose between losing the additional subsidies or making concessions to Hamas that it previously would not have made. Of course, the caveats at the end of the post apply here — if Israel doesn’t actually believe that the US would cease providing additional subsidies should another conflict breakout, then their incentives won’t actually have changed. And I confess I am a little concerned about that. But at least *in principle*, it is possible for subsidies to bring peace even if they are only provided to one side.

      • Jason November 25, 2012 at 2:57 am #

        I assume that one-sided subsidy would bring peace in a more effective way if the said subsidy is provided to the “stronger party” in the conflict dyad? For instance, if Hamas is given all the subsidy benefits, would that intervention necessarily convince Israel to abide to the conditions of ceasefire? If Tibet gets all the additional benefits of acting peaceful, why would China honor that “peace”? Here, I mean “stronger party” in the sense that both Israel and China seemingly has overwhelming advantage in military power.

        One quote that I think will be relevant: “What we call peace is often really détente.” – Venkatesh Rao,

        • Phil November 25, 2012 at 8:39 am #

          That certainly could be true empirically, Jason. I’m more than happy to admit that our model, like all models, is a simplification. But within the context of the model, at least, it doesn’t matter which side gets the subsidies. Either way, subsidies increase the surplus that’s available from peace, and that creates a material incentive to reach an agreement.

          Another way to think about it is this — regardless of who the subsides are given to, the powerful actor is impose the most favorable outcome (from their perspective) that they can on the weak actor. The more money the weak actor gets if they remain at peace, the more land or policy they will be willing to give up peacefully rather than go to war. In the end, the powerful state is better off one way or another. Either they get the subsidies, or they get more of the contested issue. Whether it always works that way in practice, I’m not sure, but that’s the intuition behind why it doesn’t matter in the model.

          Nice quote. Very much agree.

  3. donna November 23, 2012 at 1:51 pm #

    i think H and all the Is world needed ton of mind .ton of culture .to use ton of money for good of tham people and country.

  4. Thomas November 23, 2012 at 3:29 pm #


    This is a really interesting paper. I would just be curious on how you think the exact mechanism operates. From reading your paper, it seems that the mechanism you argue is that subsidies from a third-party (C) may be able to stave off conflict in the short-term (i.e. resolve the power shift problem) or in the long-term by changing the costs of conflict (“buying peace”).

    However, let me suggest an alternative (or more nuanced) mechanism, one that is more related to domestic politics, and is especially applicable in the case of Israel-Hamas, or even Israel-Egypt. I would argue the relevant story of fighting in the Middle East is the Kydd and Walter (2002) “Sabotaging the Peace.” More extremist factions have incentives to foment mistrust and spoil peace efforts, in order to gain popular support. We can imagine that third-party aid flows are actually payments for moderates to reign in extremists ( Netanyahu to reign in Likudnik hawks and Morsi the more hard-line members of the Muslim Brotherhood). In essence, aid serves a mechanism to incentivize moderates to go after/or marginalize extremist actors who would destabilize peace (“buying the coalition”). This domestic politics story is observationally equivalent to your story with your current empirics. However, perhaps you could add domestic political indicators to try to parse out whether it is a pure “buying the peace” vs. “buying the coalition.” Thoughts?

    I know this may be a bit beyond the purview of the current paper, but this paper by social psychologists Scott Atran and Jeremy Ginges may be of relevance for looking at the unintended consequences of trying to buy off hard-liners .

    Again, really cool and interesting work and would love to hear your thoughts.


    • Phil November 23, 2012 at 4:07 pm #

      Thanks for the feedback, Thomas. Excellent points.

      You are right that the mechanism described in the paper is simply increasing the effective cost of conflict. I agree that there are essentially costs for accepting agreements, and that these costs arise due to domestic processes. And if such costs exist, the real impact of subsidies has less to do with increasing the *willingness* of each side’s leadership to compromise than it does their *ability* to do so, as they will be able to use the additional funds to compensate those elements of their coalition who would be least inclined to accept the agreement. As you point out, these two stories are observationally equivalent with respect to our empirics, though the theoretical model would need to be extended in order to incorporate such dynamics. The model already has a lot of moving parts, so I’m reluctant to extend it in such a way, especially since it seems unlikely that the primary substantive conclusions would change, but I do think that your interpretation is probably closer to correct.

      The paper on compromising over sacred issues looks interesting. I will have to give it a closer read. My initial reaction is that the argument is plausible, but I’m not sure I would expect it to apply to Israel and Egypt or the current ceasefire between Israel and Hamas. No one ever said the Sinai was sacred, and the ceasefire between Israel and Hamas didn’t require either party to compromise over Jerusalem or the right of return or any of the other issues of symbolic importance. Still, an interesting argument. I will think more about this. Thanks for the link!


  5. Mark November 23, 2012 at 5:47 pm #

    Phil – I’m getting more confused about your proposal the more I read your answers – at least as it applies to the Israel – Hamas conflict. What are the concessions that Israel could be induced to make that would be appealing to Hamas?

    • Phil November 23, 2012 at 6:19 pm #

      The most obvious would be relaxing restrictions on movement in the buffer zone, allowing Palestinians to fish more than 3 miles from the coastline. It appears that Hamas leaders have already told Palestinians that they won such concessions from Israel, though the Israeli government says nothing has yet been agreed to. (See here:

      But if you think I’ve made a proposal, then I would suggest that you’ve misread my post. The point was not to make policy recommendations, but discuss the prospects of the ceasefire that has been implemented. Whether Israel *should* in some normative sense do this, that, or the other thing, or whether Hamas *should* behave a certain way, is very much outside the scope of what I’m trying to say here.

  6. Phil November 23, 2012 at 6:20 pm #

    There should be an “and” instead of a comma in that first sentence, btw. Please forgive the typo.

  7. lilojmayo November 23, 2012 at 9:56 pm #


    • Phil November 23, 2012 at 10:24 pm #


  8. LFC November 24, 2012 at 10:21 am #

    From the OP:

    In the second half of our paper, we argue that the reason why Israel and Egypt have seen nearly 40 years of peace, after fighting five interstate wars between 1948 and 1973, has less to do with Camp David and more to do with the dramatic increase in foreign aid outlays that began almost immediately after the Yom Kippur War.

    I was under the impression, but perhaps I am wrong, that the large annual US aid payments to Egypt (roughly on average $1.3 billion per year in recent years, at least that’s the figure one hears tossed around in the media) started not after the Yom Kippur War but rather after the Camp David Accords, and were indeed one reason Carter was able to get Sadat’s agreement to the treaty. But as I say, I may be wrong about this. (And if I am wrong about this, it might raise the question of why this misleading account of the history of US aid to Egypt is ‘in the air,’ or at least sufficiently in the air for me to have breathed it in.)

    Re Thomas’s suggestion above that “We can imagine that third-party aid flows are actually payments for moderates to rein in extremists (Netanyahu to reign in Likudnik hawks and Morsi the more hard-line members of the Muslim Brotherhood),” this may, I suppose, have some current applicability, but I am not convinced that a reining-in-extremists story explains the high US aid flows to Egypt and Israel since ’79, ’73 or whenever. The US may have increased aid, as the post says, in those years when the risk of Egypt-Israel conflict seemed higher, but in order to support Thomas’s suggestion it would help to show that the US increased aid in those years when ‘extremists’ seemed to be gaining ground on ‘moderates’ domestically in both countries. Short of such evidence, I would be somewhat skeptical of Thomas’s suggestion.

    And I would suggest further that US aid to Israel specifically is driven in very large part by domestic political considerations in the US, and while it may have gone up and down over the years, it is never going to dip below a fairly high threshold — regardless of what Israeli govt policy is. So, for ex, the Obama admin early on pressed Netanyahu to stop new settlements but didn’t actually threaten to try to cut aid if he didn’t agree. Bush 41 yanked some loan guarantees on one occasion in an effort to influence Israeli policy, iirc, but that’s just about the only occasion offhand I can think of where the US govt tried to use financial leverage to influence Israeli govt policy (I stand open to correction on this if wrong).

    P.s. I also don’t esp. like Thomas’s use of the terms ‘moderate’ and ‘extremist’ here: everything is relative, of course, but if your terminology leads you to label Netanyahu a ‘moderate’, there may be something wrong w/ the labeling.

    • Phil November 24, 2012 at 1:08 pm #

      Hi LFC. Good points.

      I think the reason people have the idea that it began with the Camp David Accords is partly grounded in fact (aid increased even further after Camp David, and 1979 stands out as the year in which the US gave the most money it has ever given to Israel and Egypt) and partly grounded in myth (historians frequently assert, often without any justification, that the importance of Camp David cannot be overstated). But our annual data make it pretty clear that the US was giving very little aid prior to 1973 and increased its provision dramatically in 1974. I think people also tend to forget that while it was in 1979 that a final treaty was signed, agreements on disengagement of forces were signed in January of 1974 and September of 1975. It’s also worth remembering that Sadat’s historic visit to Israel, and his speech on the floor of the Knesset, came in 1977. By the time Carter took office, the peace process was already well underway.

      I like your suggestion that one way to distinguish Thomas’ suggestion from the story we tell in our paper would be to see if the amount of aid given in any year corresponds to observable fluctuation in the influence of domestic factions opposed to peace. That’s a great point. I’ll have to think more about how to do that.

      I think you’re exactly right about the importance of domestic politics in explaining US provision of foreign aid to Israel. I tried to make clear in the post that I think our story has more to say about the aid given to Egypt than to Israel, but I probably should have elaborated on this more.

      • Phil November 24, 2012 at 1:14 pm #

        To be more precise: the average amount of economic aid provided to Israel and Egypt (combined) from 1948 to 1973 was just over $400 million. The average amount of economic aid provided to Israel and Egypt (combined) from 1974 to 1977 was just over $3 billion. From 1978 to 1982 was a little over $4 billion. The average since then has been $2.5 billion, though as you note it has dropped below $1.5 billion in recent years.

      • LFC November 24, 2012 at 1:45 pm #

        Phil, thanks for the reply and for the reminder that there was a long lead-up to the treaty.

        • LFC November 24, 2012 at 1:53 pm #

          Re the aid figures: I’m assuming the definition of “economic aid” there is fairly elastic such that a lot of it is actually security-related.

          • Phil November 24, 2012 at 2:26 pm #

            It’s a pretty loose term, but there is a separate category for “military” aid. In previous drafts of the paper, my co-author and I lumped the two together (and got the same results). But some people raised concerns about whether military aid is given for the same reasons as economic aid, and the fact that our model assumes subsidies are a pure consumption good rather than something that influences future military capabilities. But in the interests of full disclosure:

            Average annual military aid to Israel and Egypt (combined), 1948 to 1973: ~$250 million
            Average annual military aid to Israel and Egypt (combined), 1974 to 1977: ~$4.5 billion
            Average annual military aid to Israel and Egypt (combined), 1978 to 1982: ~$5.5 billion
            Average annual military aid to Israel and Egypt (combined), 1982 to 2007: ~$4.5 billion

    • Thomas November 24, 2012 at 6:53 pm #


      Good points all around. LFC, I think that looking at yearly shifts in domestic political coalitions and how that affects aid flows may be fruitful. However, we know that domestic political factors (especially in Israel) are a function of the foreign policy environment, so teasing out the effects of purely domestic politics vs. external increases in the threat of conflict may be difficult.

      I have two quibbles with LFC’s comments.

      1) I recognize that ‘moderate’ and ‘extremist’ are indeed both relative and normative….and that we need to be careful between conflating the two. But with respect to Israeli politics and especially on the Gaza Conflict, Netanyahu is much more moderate then most members of his coalition and many other, more centerist opposition parties who wanted to invade Gaza This is not to say Netanyahu is a moderate in the normative sense of the word. Yet, there are policy insights to be gained from a fuzzy rank-ordering of the Israeli political landscape.

      2) I agree that on average aid to Israel is fairly stable (how much this is do to domestic politics versus bureaucratic inertia is open to debate). However, I think the recent Hamas-Israel hostilities, and Israel’s decision to avoid a ground invasion and abide by the cease-fire was largely a result of financial/diplomatic promises by Obama over the Iron Dome . This fits Phil’s story quite nicely.

      In fact, we can argue that there are indeed perverse incentives at play for the US vis a vis the Israelis. It could be perhaps that Israelis are/were more willing to initiate a dispute with Hamas partially in order to extract concessions from the US on core issues they are care about (Palestinian statehood at the UN, Iran, Missile Defense). Phil, any thoughts?

      • Phil November 25, 2012 at 7:44 pm #

        That’s a really interesting idea, Thomas. I’ve heard it kicked around before, and it strikes me as plausible. I’m not sure how big a factor it might be, but I can definitely see it being part of the story.

        I question the extent to which that motivated Israel’s behavior because the 8 day limited conflict cost at least 5% of Israel’s annual defense budget ( I don’t think the concessions won from the US are likely to do much to offset that.

        Similar arguments have been made about perverse incentives with counterterrorism aid — that countries who receive such aid have an interest in continuing to have a terrorism problem. Andrew Boutton at Penn State is doing some interesting work on this.

  9. Kyle November 26, 2012 at 2:58 pm #

    I like the post and the ideas in the working paper. As you mention, the big question about the effectiveness of such subsidies is whether or not they can be credibly sustained in the long run. On that, I wonder if you have run your models as a function of lagged aid differences and not just levels. I buy the link between aid levels and peace, but I’d be curious to see if short-term bumps in aid do not well correlate with peace and may actually destabilize post-conflict peace. I’m not sure how to best capture the measure of short-term bumpiness (maybe an interaction between like the 5th difference and the 1st difference, which would allow you to see if aid is higher than it was 5 years ago but less than it was in the previous year), but this could produce some additional tests of the various arguments about outside pressure and peace.

    • Phil November 26, 2012 at 5:54 pm #


      This is an interesting suggestion. I just reran the models, replacing the level of aid (and its interactions with parity and shifts in power) with the annual change in aid (and appropriate interactions) and then again with both. In neither case was there any evidence that the year-over-year change in aid had a significant impact in the predicted number of forceful MIDs. Of course, we need to be careful in how we interpret this, since the average annual change in aid is very small. It could be that there are real effects here that we’re just not able to pick up because there is too little variation in the behavior that occurred historically for us to gain any leverage on what might have occurred had history unfolded differently.