Problems and prospects for democratic settlements:
South Africa as a model for the Middle East and Northern Ireland?
Courtney Jung, Ellen Lust-Okar, and Ian Shapiro
February 26, 2004
In the 1970s South Africa, Northern Ireland, and the Arab-Israeli conflict were often grouped together
as the world’s most intractable political conflicts. Histories of intense ethnic, racial, and religious
conflict had led generations of commentators to regard them as poor candidates for democracy.
Curiously, the conflicts have diverged markedly in subsequent decades. South Africa moved through
an unexpected transition to majority rule, while negotiations in the Middle East and Northern Ireland
have moved forward dramatically at various times and then stalled. We argue here that the old
conventional wisdom was wrong—blinkered by the misleading literature on “divided societies.”
Instead, these conflicts are best seen as a distinctive type of democratic transition in which achieving
an enforceable settlement depends on the democratic legitimacy of the negotiating principals. Their
need for democratic legitimacy creates constraints and also presents opportunities. We show how the
principals navigated the constraints and took advantage of the opportunities in South Africa, how they
have failed to do this in the other two other two conflicts thus far, and what would need to happen for
them to succeed in the future.
In the 1970s, the political conflicts in South Africa, Northern Ireland, and the Middle East
were often grouped together as among the world’s most intractable. They exhibited profound racial
and ethnic animosities, reinforced by linguistic, cultural, economic and religious differences, and
solidified by decades of more-or-less violent confrontation. They were often held out as paradigm
cases of “divided” societies, and there seemed little chance of a transition to peaceful, let alone fully
democratic, arrangements in any of them. Whether one focused on the players contending for power,
the histories of the conflicts, or the capacities of outsiders to influence events, the prospects seemed
dim for negotiated settlements.
The conflicts have diverged remarkably in subsequent decades. South Africa, often depicted in
the grim 1970s as the most intractable of intractables, moved through a comparatively peaceful four-
year transition to majority rule in a unitary state. Democratic elections in 1994 and 1999 put the
African National Congress (ANC) securely in power without civil war, economic collapse, or
catastrophic white exodus. To be sure, the continuing economic and social challenges are enormous,
with a third of the population unemployed and one in nine infected by the HIV virus, but by most
measures South Africa has weathered the transition well. Democracy may not yet be entrenched in
South African politics, but it seems at least to have a fighting chance.
Northern Ireland has also made important advances since negotiations began in earnest in
1996. Both Republicans and Loyalists committed to cease fires that have held, and most serious
violence has abated sufficiently that people have started to think peace a realistic possibility. The two
sides signed an agreement in 1998 that majorities of both Catholics and Protestants supported. Yet the
future of the 1998 Good Friday agreement remained precarious, at best, in 2004. The failure of the
power-sharing government to work had led to its repeated suspension by Westminster, and eventually
the re-imposition of direct rule in October 2002. Whether the paramilitary groups would disband and
the Executive and Assembly would be revived remained to be seen.
Establishing peace between the Palestinians and Israelis has been even more elusive. There
have been some major turning points in the Arab-Israeli conflict and periods of great optimism, most
notably following the negotiation of the Camp David Accords and subsequent signing of the Israeli-
Egyptian peace treaty in 1979 and the Palestinian-Israeli negotiation of the Oslo Accords in 1993.
There have also been numerous less dramatic “openings” from one side or another. PLO acceptance of
a two-state solution in 1988, Syria’s decision to support the multilateral Madrid Conference in 1991,
and Israel’s decision to negotiate directly with the PLO all provided windows of opportunity for their
negotiating partners. However, they have often been either unable or unwilling to seize the
opportunities that emerge when one side makes concessions. For instance, in 1998 Benjamin
Netanyahu rendered Yasir Arafat’s concessions in the Wye Accords useless when he unilaterally
suspended implementation, and Arafat refused Ehud Barak’s concessions at Camp David II.
Backtracking and disappointing failure have been so frequent that the peace process often
seems ritualistic and pointless. Different leaders participate more or less grudgingly at different times,
almost always under intense American pressure. This was manifest in President George W. Bush’s
“road map” for peace in the summer of 2003. Israeli Prime Minister Ariel Sharon was induced to use
the word “occupation” for the first time in relation to the conflict, but his insistence that this referred to
people and population centers, not the West Bank as such, made the seeming concession nearly
meaningless. The concessionary speech delivered by Palestinian Prime Minister Mahmood Abbas at
the Aqaba summit, widely reported to have been drafted by the White House, led Hamas to break off
talks with the Palestinian Authority on ending the violence, and further depletion in his already
negligible approval ratings among Palestinians.
It was unsurprising that the unilaterally declared
“truce” by Hamas and other militant groups quickly collapsed in light of the vast differences that
remained over territory, settlements, “the right of return,” disarmament, and prisoners, and the dearth
of popular support on both sides for politicians showing any inclination to bridge these gulfs. The
Middle East peace process continues to vindicate the 1970s diagnosis by going nowhere—if often by
Byzantine routes at enormous human and economic cost.
Whence these divergent outcomes? Are there lessons to be drawn from South Africa’s
comparative success that can illuminate the ongoing dynamics in Northern Ireland and the Middle
East—both why they have not yet succeeded and whether they can succeed in the future? Perhaps the
answer is no. The issues, stakes, and constraints might all be so different that each situation follows
an independent logic. Indeed, some might object that the three situations are not comparable. After all,
South Africa involved a transition to majority rule within a unitary state whereas both the Middle East
and Northern Ireland involve creating or maintaining partitions with power-sharing. But this gets the
cart before the horse. In fact, stakeholders in all three of the conflicts have entertained variations on
each of these solutions at different times. Ironically, partition might reasonably have been judged more
likely in South Africa, ex ante, than in the other two conflicts—if only for demographic reasons.
See “Sharon backtracks on Israeli ‘Occupation’,” Newsmax.com Wires (May 28, 2003) at
http://www.newsmax.com/archives/articles/2003/5/28/100340.shtml (6/13/2003) and James Bennett, “Hamas breaks of talks
on stopping attacks on Israel,” New York Times, June 11, 2003: A1, A7.
Others agree with us that these three cases exhibit important similarities, but they misconstrue
the basis for comparison. These scholars focus on the putatively “divided” character of the three
societies—suggesting that conflicts engendered by ethnic, racial, and religious divisions preclude
Hence the dire predictions about South Africa in the 1960s, ’70s and ’80s. In our
account, negotiations such settings are best seen as a distinctive class of transplacements or negotiated
What makes them distinctive is not their “divided” character, but rather that they are
imperfectly democratic. The ancién regime
is a flawed democracy, not a conventional authoritarian
The need for democratic legitimacy provides the impetus for regime change, shapes the
negotiations in distinctive ways, and structures the available settlements. Success depends not on the
key players finding the holy grail, but rather on them converging on a solution that their constituents
will accept as legitimate.
It will be clear from this claim that ours is not a deterministic argument. Individual players
make decisions that could be made differently, often with consequential results. Moreover, many
contingencies affect the outcomes of negotiations. Had F.W. De Klerk been shot by a disgruntled
right-winger before the 1992 referendum, the South African transition might have fallen apart. Had
Yitzhak Rabin not been shot in 1995, a successful agreement between Israelis and Palestinians might
by now have been concluded and implemented. We might then be trying to explain Middle East
success in contrast to South African failure.
This is not to say that resolution is exclusively dependent
See Colin Knox and Paul Quirk eds., Peace building in Northern Ireland, Israel and South Africa: Transition,
Transformation and Reconciliation (Basingstoke, Eng.: Palgrave, 2000); Ian Lustick, “Necessary risks: Lessons for the
Israeli-Palestinian peace process from Ireland and Algeria. Middle East Policy, Vol. I, No. 2 (1994), Ian Lustick, Unsettled
states, Disputed Lands: Britain and Ireland, France and Algeria, Israel, and the West Bank-Gaza, (Ithaca: Cornell University
Press, 1994); Donald Akenson, God’s Peoples: Covenant and Land in South Africa, Israel, and Ulster (Ithaca: Cornell
University Press, 1992); Heribert Adam ed. Comparing Israel and South Africa: Prospects for Conflict Resolution in Ethnic
States, (Colorado: Univ. Press of Colorado, 1990); Herman Giliomee and J. Gagiano J., The Elusive Search for Peace. South
Africa, Israel, Northern Ireland, (Cape Town: Oxford University Press, 1990); Benjamin Gidron, Stanley Nider Katz,
Yeheskel Hasenfeld, eds. Mobilizing for Peace: Conflict Resolution in Northern Ireland, Israel/Palestine, and South Africa
(New York: Oxford University Press, 2002).
So for example we rely on the work of Guillermo O’Donnell and Philippe Schmitter, Transitions From Authoritarian
Rule: Comparative Perspectives (Baltimore: Johns Hopkins University Press, 1986), Adam Przeworski, Democracy and the
Market (New York: Cambridge University Press, 1991) and Samuel Huntington, The Third Wave (Oklahoma: University of
Oklahoma Press, 1991) to think through the logic of SAMENI conflict resolution. These three negotiations have rarely been
analyzed in the transplacement literature. One notable exception is Timothy D. Sisk, Democratization in South Africa: The
Elusive Social Contract (Princeton: Princeton University Press, 1995). In Jung and Shapiro, “South Africa’s negotiated
transition,” we also used the model of transplacements to think through the dynamics of negotiation in the South African
Counterfactual speculation is inherently difficult, though we adduce considerable evidence in support of these claims
below. With respect to South Africa it merits reporting that F.W De Klerk believes that had he been assassinated after the
March 1992 referendum the negotiations would likely have been concluded succesfully, but that this is much more doubtful
had it occurred before the referendum when the government was losing by-elections to conservatives. De Klerk reports that
the decision to call the referendum was the only unilateral decision of his presidency. He consulted no one in the cabinet or
on luck and contingencies of leadership—defying the possibility of useful theory. Such contingencies
make it impossible to predict success in any given instance, but this does not exhaust the theoretical
agenda. For one thing, on our account it is possible in many situations to predict failure, and, perhaps
more important, to say something about how those situations would have to change for success to
become a possibility. For another, when success is possible, we can and do develop accounts of the
conditions under which it becomes more or less likely.
Our procedure is as follows. We begin, in §1, by describing how the negotiations under study
constitute a distinctive class of transplacements. We label these conflicts with the compound acronym
SAMENI, to signal that ours is an inductive effort at generating hypotheses—based on the South
African experience and a close examination of how similar dynamics have played out in the Middle
East and Northern Ireland. In §2 and 3 we explore the conditions that facilitate the initiation,
negotiation, and consummation of SAMENI agreements, explaining how circumstances coincided to
allow South Africa to overcome the barriers to democratic legitimation, moving through all three
stages with comparative ease, while negotiations in the Middle East and Northern Ireland have
continually snagged. This leads to a discussion, in §4, of how negotiations could have succeeded in the
Middle East and Northern Ireland, and the conditions under which they might do so in the future.
1. The character of SAMENI negotiations
SAMENI negotiations resemble other transplacements in three ways. First, because they
concern political fundamentals, the stakes are inevitably high. Questions of sovereignty, involving
regime type, territorial boundaries, or both, are at issue. If an agreement is reached and implemented, it
will lead to irreversible changes in a major part of political reality. The negotiations involve
intertwined issues of personal security, economic survival, and collective destiny that have often been
politicized by decades of conflict. Even if negotiations fail, or the agreement is not implemented, the
power balance is likely to change, making return to the status-quo ante difficult or impossible. Political
futures are on the line for the principals, giving them large and increasing stakes in the outcomes. In
the National Party leadership because he knew they would oppose it. Even if a new leader wanted to call a referendum, he
doubts that either of the likely contenders (Pik Botha and Roelf Meyer) would have been able to do so, given the need to
establish themselves in the party leadership. At the very least the process would have been significantly delayed (interview
with author, December 9, 2003). Given our discussion of the importance of timing below, this might well have been
sufficient to derail it permanently.
If de Klerk had been killed, the only conceivable replacement who might have been able to carry the NP and the
military through negotiations was Roelf Meyer. But Meyer was a very junior minister, without much standing in the party. It
seems more likely that Pik Botha would have assumed leadership and resorted to reforming apartheid. SADF support for
the transition was at best tenuous at the time of the referendum.
short, like other transplacements, SAMENI negotiations exhibit the life-or-death quality of politics that
is about the basic rules of the game.
This is why transplacement negotiations are so fragile. Reformers and moderates are still, in
most ways, adversaries who must constantly judge one another’s agendas and abilities, as well as
reassess their own. They can signal to each other their intention to continue in the process, but to do
so they must take decisive steps in facing down domestic opposition even before it is clear that an
enforceable agreement will be reached. As a result, although the principals know that success may
write them into the history books, the risks are huge. At critical junctures they must be willing to face
down historical allies on their own flanks to gain a prize that will be theirs only if their negotiating
adversaries can do the same thing. Moreover, they have little reason to trust those with whom they are
dealing. It is, in short, one thing for there to be a potential coalition in favor of a negotiated settlement;
quite another for it to form and sustain itself long enough to get the job done. Because this requires
splintering existing coalitions and fending off attacks from historical allies who feel threatened or even
betrayed, it takes creative ingenuity, courage, and luck.
Second, like other transplacements SAMENI negotiations occur in a power stalemate in which
no one can impose change. Neither the regime nor its opponents can dictate a solution, yet there is a
potential coalition of government reformers and opposition moderates who may be able to negotiate an
agreement both prefer to the status quo.
For transplacements to succeed, the innovative coalition must
remain sufficiently strong that, should an agreement be reached, its members can carry their
constituencies along and impose the settlement on government hardliners and opposition radicals who
resist it. Moreover, they must do so at the same time. Because multiple factors must come together in
the right sequences, there are many more ways for all negotiated settlements to fall apart than for them
A third way in which SAMENI negotiations resemble other transplacements often goes
unnoticed due to the widespread proclivity to hive them off as “divided societies.” Viewed in that way
the conflicts seem to involve particularly intractable forms of political violence. It is true that there
have been periods of considerable violence in South Africa, the Middle East, and Northern Ireland, but
this scarcely differentiates them from other transplacements—as in Chile in the 1970s or El Salvador
in the 1980s for example. Yet it is not so much the
of violence that commentators focus on as
The language of hardliners and reformers, moderates and radicals, is used in the literature on democratic transitions.
See in particular Przeworski, Democracy and the Market, pp.67-70 and Huntington, The Third Wave, pp.151-164. The terms
, and in particular the fact that it occurs among groups that define themselves by reference to
such categories as race, religion, or ethnicity. The common assumption that these categories are
ascriptive, if not primordial, leads people to misidentify the conflicts as inherently zero-sum, and to
miss the possibilities for unanticipated alliances and the redefinition of political identities as
negotiations evolve. That there was a non-Solomonic settlement in South Africa surprised many
people. If our analysis is correct they should not have been surprised, and those who continue to insist
on the sui generis character of the violence in the Middle East and Northern Ireland should not make
the same error.
Yet on our account SAMENI transplacements are nonetheless distinctive. They are a type of
transplacement that is both complicated and motivated by conditions of imperfect democracy. Unlike
standard transplacements in countries like Spain, Poland, and Chile, the government is democratically
elected. Yet they are imperfectly democratic because large populations under the government’s control
are disenfranchised or partly enfranchised in ways that are widely seen as unjust. This reality gives the
regimes inherent legitimacy problems because they must claim to be democratic when they obviously
are not. By entering negotiations, reformers acknowledge, however implicitly, this deficiency in their
system. This means that they are usually on the defensive—arguing about the terms and pace of
change rather than its necessity. This in turn means that no settlement can succeed unless there is broad
agreement that the democratic deficit that gave impetus to negotiations has been substantially
attenuated, if not abolished. Once parties to a conflict appeal to democracy as their source of
legitimation, widely accepted democratic norms rule out racial oligarchies, and in today’s world they
even make religious and ethnic oligarchies suspect. Moshe Halbertal explained this imperative well
following the collapse of the July 2000 Palestinian-Israeli Washington summit: “Between the
Mediterranean and the Jordan there are roughly five million Jews and five million Palestinian Arabs.
You cannot have a Jewish and democratic state without dividing this land, and those who oppose that
are dooming Israel to an apartheid state, which might have secure borders, but might not be worth
securing. Barak is on a purely Zionist mission to bring Israel back into borders where it can be Jewish,
just, democratic—and secure.”
The opposition in SAMENI conflicts often emanates from a liberation movement and is not, as
such, democratically elected, but it gains significant leverage from the fact that the government lacks
reformers and moderates do not refer to the content of political ideologies. Rather, they denote players who are willing to
entertain outcomes that differ from their political ideals in search of a mutually acceptable solution.
Moshe Halbertal, as quoted by Thomas Friedman, “Yasir Arafat’s Moment,” New York Times, July 28, 2000: A21.
democratic legitimacy. Yet by entering negotiations, the opposition inevitably becomes democratically
constrained as well. Its leaders must be able to claim credibly that they represent a major constituency,
if not the majority, and to move toward a settlement that will be popularly validated. In short, although
the regime and its opponents may both be imperfectly democratic, they claim to be democrats and
depend on popular support in a more robust sense than the players in other transplacements. The need
for democratic legitimation greatly complicates negotiations, defying attempts to reduce them to
stylized elite games. We are thus sympathetic to Elisabeth Wood’s contention that the transitions
literature has been overly focused on elite interactions, with insufficient attention to the larger political
contexts within which they occur.
But where Wood contends that such negotiations are driven from
below, we take a more interactive view. Negotiators are constrained by popular opinion, but to succeed
elites must make the right choices at critical junctures—including choices about how to respond to
popular opinion and when to try to shape it. One of the trickiest problems arises from the reality that
negotiating a settlement usually involves concessions that force the principals to move away from their
mandates. The challenge then becomes finding ways to avoid alienating constituencies whose
endorsement is essential to the settlement’s legitimacy.
The central question in all transplacements is: can the reformers and moderates agree on a
settlement and successfully face down the hardliners and radicals on their flanks? However, SAMENI
transplacements are distinctive in that the parties must also maintain enough grass roots support that
backers of the
continue to see the settlement as legitimate while partisans of the new
dispensation regard it as repairing the democratic deficit. Moreover, the dynamics of negotiations will
be affected by democratic turnovers in power. If the negotiating government falls at the polls, as has
often been the case in Israel, new players must then establish their credentials as bona fide reformers
intent on concluding an agreement. As well as constantly reassessing their own interests in proceeding,
both sides must thus worry about whether the other can maintain enough support among their core
constituencies to carry through their own side of the bargain. They must also worry about how
concessions they might make threaten to alienate their own supporters. Barak’s willingness to put
sovereignty over parts of Jerusalem on the table for the first time in the failed 2000 Washington
negotiations is a case in point. His subsequent loss of popularity at home was partly linked to the
Elisabeth Wood, Forging Democracy From Below: Negotiated Transitions in El Salvador and South Africa
(Cambridge University Press, 2000).
realization that it would be difficult, if not impossible, for any Israeli leader to declare Jerusalem off
limits in future negotiations.
2. Onset of SAMENI negotiations
Catalysts for SAMENI negotiations can take the form of sticks, carrots, or—more likely—
both. For government reformers the main stick will likely be an increasingly costly, deteriorating
status quo, depleting their political capital and increasing their will to negotiate. This may be because
of internal developments such as terrorist bombings or an ungovernability campaign; or because of
external factors such as sanctions, pressure from international human rights groups or a powerful ally.
Changing structural, global, or popular constraints may render the status quo less viable, and
alternatives more readily imaginable. Evolving ideological paradigms can also shift perceptions of the
viability or meaning of persisting in conflict. For instance, the fall of communism or the increasing
bankruptcy of race as an organizing principle of political and social life might undercut the grounds
that have justified violence hitherto. Carrots could include the prospect of peace and an end to pariah
status in world opinion, a variety of economic incentives, or a desire to do the right thing and go down
in history as a statesman.
Comparable considerations apply to the opposition. Sticks might include the inability to sustain
grassroots support for a costly and unwinnable guerrilla war, international pressure, depleted weapons,
or dissension within the liberation movement. Among the carrots may be the legitimation afforded by
recognition and talks with the government, the allure of power, access to international players,
promises of economic support from third parties, or the advantages of peace and prosperity. In all three
cases under discussion, the combination of sticks and carrots ushered in unprecedented negotiations
that held out the hope of ending decades of intractable conflict.
2.1 How the unthinkable became thinkable in South Africa
Throughout the 1980s the South African government faced a deteriorating status quo. The
ungovernability campaign mounted by the United Democratic Front (UDF) massively raised the costs
of keeping order in the townships by organizing a generation of young black activists with a more
A poll conducted by the Palestinian Center for Policy and Survey Research (PCPSR) in Ramallah and the Harry S.
Truman Research Institute for the Advancement of Peace at the Hebrew University, Jerusalem found that 55.7 percent of
Israelis surveyed believed that “Israel made too much of a compromise” at the Camp David Summit. PSR, “Results of
Israeli Public Opinion Poll: July 27-31, 2000,” www.pcpsr.org/survey/polls/2002/p1israelipoll.html (6/17/03).
Some incentives are long-standing. For example, most observers argue that business elites in South Africa, Northern
Ireland, and Israel have long favored a peace settlement, and we have found striking evidence of this in Israel, where all 15
of 15 top business executives we surveyed favored an agreement. However, there is no evidence in any of these cases that
the pressure from the business community was responsible for either beginning or continuing negotiations.
militant opposition style than their parents.
The currency collapse that followed South Africa’s
inability to meet international debt obligations in 1985 sent the economy into a tail-spin, and the
relentless chorus of outside political and economic pressure began to be matched by attacks on
apartheid from growing numbers of Afrikaner intellectuals.
By the second half of the 1980s, polls
revealed that most whites believed that apartheid threatened the country’s future.
NP confidence in
the medium-term viability of the apartheid state was particularly shaken by the escalation of violence
following the collapse of the second phase of CODESA roundtable negotiations in May 1992.
The two most important carrots had to do with the collapsing Soviet Empire after the mid-
1980s. Because the leaderships of the ANC and the South African Communist Party (SACP)
overlapped substantially, white fear of majority rule was conflated with fear of communism. After
1989 a communist government in South Africa was no longer a serious threat, and white elites began
to realize that majority rule need not mean the destruction of capitalism or the expropriation of private
The political corollary of this was a more flexible and pragmatic ANC leadership with
whom serious negotiations, if not yet partnership, could be considered. That this would also mean an
end to pariah status, the possibility of economic revival, unfettered overseas travel, and countless other
benefits of normalization no doubt also helped. Once white South Africans began thinking the
unthinkable, it could start looking attractive.
NP carrots were ANC sticks. In the early 1980s a strapped USSR had stopped ANC and SACP
financial backing and military training on the grounds that South Africa was not in a revolutionary
This was a closely held secret within the ANC leadership, but it was only a matter of time
The UDF was the backbone of internal opposition to apartheid in the 1980s, and widely considered the internal wing
of the then-banned ANC. Jeremy Seekings, The UDF: The United Democratic Front in South Africa, 1983-1991 (Athens,
Ohio: Ohio University Press, 2000).
An anti-Apartheid Afrikaner intellectual movement had its foundation in the late 1960s. Heribert Adam and
Hermann Giliomee, Ethnic Power Mobilized: Can South Africa Change (New Haven and London: Yale University press,
Suzanne Booysen, “The Legacy of Ideological Control: The Afrikaner Youth’s Manipulated Political
Consciousness,” Politikon, 16, no.1 (1989), pp. 7-25
The acronym stands for Conference on a Democratic South Africa. These were part of the “prenegotiations” in that
the government insisted in involving over twenty interests, including those who had no interest in a democratic transition
such as the extreme white right and the Inkatha Freedom Party. Predictably they went nowhere. See Courtney Jung and Ian
Shapiro, “South Africa’s Negotiated Transition: Democracy, Opposition, and the New Constitutional Order.” Politics and
Society Vol. 23, No 3 (1995), pp. 285-286.
In fact white economic elites realized this before the politicians, and began secret talks with the ANC as early as
1986, in Lusaka and Dakar. See Timothy D. Sisk, Democratization in South Africa: The Elusive Social Contract (Princeton:
Princeton University press, 1995), p. 78.
“Statement of the National Executive Committee of the African National Congress on the Question of Negotiations”
Lusaka, 9 October, 1987, cited in Steven Friedman, The Long Journey: South Africa’s Quest for a Negotiated Settlement
(Johannesburg: Raven Press, 1993), pp. 10-11.
until the government would know it as well. Overtures to China led nowhere, forcing the ANC-in-exile
to rethink its military strategy and start building up internal opposition. The UDF was formed in 1983
in opposition to Tricameral Parliament elections which had offered some representation to “coloured”
and other disenfranchised racial minorities. This ushered in a new era of populist opposition with a
massive ungovernability campaign in the townships.
Widespread internal unrest raised the stakes for
the NP government at the same time as it marginalized the Pan Africanist Congress (PAC) and Inkatha
forces within the liberation movement. But the government’s ferocious repression of the opposition,
particularly after 1986, as well as its huge reserves of military and paramilitary power (South Africa
had—and has—by far the most powerful and best equipped army in Africa), made it plain that the
Soviets were right about the low odds of successful revolution.
The most important carrot in getting the ANC leadership to the bargaining table was the
prospect of power. The low odds of military success meant that the De Klerk government’s
willingness to negotiate had to be taken seriously. Even if the ANC had to accept the prospect of
power sharing in an interim government at least, this was surely better than nothing. Moreover, it
opened up the possibility of a new status quo that could subsequently develop into full majority rule—
as turned out to be the case.
2.2 Shifting constraints and possibilities in the Middle East
The Israeli decision to enter negotiations mirrored the South African one in several ways.
Maintaining the status quo had became more expensive with the eruption of the first intifada. The
anger of Palestinian youth, and organizational efforts by the Unified National Leadership of the
Uprising (UNLU), took a toll on Israeli public support for the occupation. Israelis increasingly
returned to their pre-1967 beliefs that occupation of the West Bank and Gaza was neither feasible nor
No doubt a drop off in world support for Israel, caused by vivid nightly depictions of the
confrontation between stone throwing youth and well-armed Israeli soldiers on CNN, reinforced this.
The Tricameral Parliament was a last-ditch attempt by the NP to refashion the tatters of apartheid’s divide-and-rule
strategy by creating separate houses of parliament for Colourds and Indians (though none for Blacks).
Israeli support for talks with the PLO consequently increased. A New York Times poll in April 1987 found that
only 42 percent of respondents favored such talks. In contrast, a New York Times poll conducted in March 1989 found 58
percent of Israelis supported negotiations with the PLO if it recognized Israel and ceased terrorist activity. A poll by Yediot
Aharonot found similar results the following month, with 59 percent of respondents supporting talks with the PLO. Mark
Tessler, A History of the Israeli-Palestinian Conflict (Bloomington: Indiana University Press, 1994): 724-725. An
independent poll conducted for PM Rabin on the eve of the Oslo Agreement also confirmed that the public would support an
agreement, even if Arafat was involved. David Makovsky, Making Peace with the PLO: The Rabin Government’s Road to
the Oslo Accord (Washington: Washington Institute for Near East Policy, 1996), p. 62.
By the early 1990s, the Israeli economy was also suffering from the influx of Soviet Jews,
economic dimension to the political malaise and prodding the leadership in the direction of
By 1991, there were increasing pressures within Israel for an agreement with the Palestinians.
After the Gulf War, which neutralized Syria and moved moderate Arab states into closer alliance with
the US, the Bush Administration decided that the time for negotiations was ripe—adding both to the
pressure and sense of opportunity for Israel.
There was also a growing domestic constituency
supporting an agreement with the Palestinians. Frustrated by the intifada and alarmed by the noticeable
decrease in US government support, army generals, the business community, and a large segment of
the Israeli political left called for a peace process.
Rabin capitalized on this frustration when he ran
successfully against Shamir, promising an “autonomy agreement” with the Palestinians and restoration
of US-Israeli relations.
In 1993, Rabin decided for the first time to talk to the PLO and Arafat, then
the most powerful Palestinian leader who would be essential to any deal.
At the same time, the costs of entering negotiations were falling for Arafat. He had weathered
the internal criticism for acknowledging the state of Israel in 1988 and, with the help of political
platforms set forth during the intifada, gained general Palestinian acceptance for a two-state solution.
He also faced less opposition from Arab leaders to entering negotiations, with Egypt’s move toward
peace in 1979, King Hussein’s relinquishing of the West Bank in1988, and Syria’s willingness to
participate in the multilateral Madrid Conference in 1991. His perception of Israel’s strength might not
have changed, but his expectations about the domestic and regional costs of negotiating were notably
lower after 1991 than at any previous time.
The intifada had strengthened the Palestinians by putting pressure on Israel to negotiate, but
the PLO also faced mounting pressure to enter negotiations. It had become increasingly isolated since
During the 1990s, 900,000 former Soviet immigrants arrived, the largest immigration to Israel since the mid-1950s.
Over 330,000 of these arrived in 1990 and 1991. Nurit Yaffee and Dorith Tal, “Immigration To Israel From The Former
Soviet Union,” 2000, p. 4. http://www.cbs.gov.il/engprfl.htm. (6/18/03).
James A. Baker III (with Thomas DeFrank), The Politics of Diplomacy: Revolution, War and Peace, 1989-1992
(New York: Putnam’s Sons, 1995), p. 422.
On the development of business community support for agreement/continuing peace process, see “In the Middle
East: ‘Peace is now irreversible,’” Business Week, November 20, 1995, pp. 62-64.
David Makovsky, Making Peace with the PLO, p. 12.
Rabin continued to oppose direct deals with the PLO until mid-1993, and only after August approved the draft Oslo
Accord. Avi Shlaim, “The Oslo Accord,” Journal of Palestine Studies, 91 ( Spring 1994), p. 32.
Pamphlets distributed during the intifada, as well as a coordinated demonstration of Palestinian, Israeli and
international actors in 1990, called for “Two States for Two Peoples,” a marked contrast from earlier calls for the
the early 1980s—abandoned by Egypt in the 1979 Camp David agreement and thrown out of Lebanon
in 1982. Yet these challenges paled by comparison with the events of 1990-91. The fall of the Soviet
Union and Eastern Bloc diminished military and financial support. Then, Arafat took the wild, and
ultimately disastrous, gamble of supporting Iraq during the Gulf War—leading the PLO to the edge of
financial and political bankruptcy.
Dismayed Gulf States withdrew financial support and Kuwait
threw Palestinian workers out of the country. Defeated Iraq was in no position to help. The PLO could
not pay monthly salaries, let alone support its functions.
Finally, Syria had seized the opportunity of
the Gulf War to move closer to the US, first agreeing to join in the US-led coalition and then accepting
an invitation to the Madrid Conference. So it is not surprising that, following the failed USSR coup
attempt in August and the Syrian acceptance of the Madrid Conference, the PLO authorized a
Palestinian delegation, led by Haider Abdul Shafi, to start negotiations. As Farouk Qaddumi explained,
it was time for the PLO to join the peace process or exit history.
Both the intifada and the presence of a non-PLO negotiating team at Madrid were creating an
alternative Palestinian leadership. Much as Arafat welcomed the new legitimacy for the idea of a
Palestinian state, it was becoming alarmingly possible for Palestinians and others to imagine this state
without a role for him and the PLO. Negotiating at Oslo was a way to preserve their role, though it
came at a price because negotiations involve concessions, and, until an agreement is actually
consummated, making concessions increases the leadership’s vulnerability to a flank attack. Fear of
losing his grip on the Palestinian leadership propelled Arafat to accept concessions, and it also forced
him to confront the possibility of an end-game rather than an endless peace process. Time and
momentum were on his side, but if things dragged on for long enough without an agreement, they
would turn against him.
2.3 Opportunities to end stalemate in Northern Ireland
Negotiations in Northern Ireland resulted from a different mix of sticks and carrots. Neither
Britain nor the Irish Republic suffered unsustainable damage from the conflict in Northern Ireland,
elimination of Israel. Hanan Ashrawi, This Side of Peace (New York: Touchstone, 1995). Mark Tessler, A History of the
Israeli-Palestinian Conflict (Bloomington: Indiana University Press, 1994).
PLO coziness with Iraq was not new. In the late 1980s, Iraq had promised it that it would commit 54 army divisions
against Israel after end of the Iran-Iraq war, and Arafat had described Iraq as defender of “eastern gate” of the Arab nation.
Similarly, in December 1989, Arafat praised the unveiling of a new Iraqi ballistic missile as a “gift to the intifadah.” Yazid
Sayigh, Armed Struggle and the Search for State: The Palestinian National Movement (Oxford: Clarendon Press, 1997), pp.
Approximately 400,000 Palestinians were thrown out of the Gulf countries, and the PLO lost an estimated $10
billion between 1991-1993. Graham Usher, Palestine in Crisis: The Struggle for Peace and Political Independence after
Oslo (London: Pluto Press, 1995.), pp. 1-2.
and the conflicting parties within Northern Ireland had ample popular legitimacy and access to the
limited type of military equipment needed to continue the conflict. The Northern Ireland economy was
depressed, but economic problems were not generally blamed on the Troubles, and budget transfers
from Britain ensured that Northern Ireland was able to spend beyond its means.
relevant parties were engaged in talks to end the conflict for much of the 1990s. Why?
Solving this puzzle requires attention to the unusual combinations of participants to the
Northern Ireland conflict. The transplacement model of hardliners, reformers, moderates, and radicals,
which captures the principal dynamics of negotiations in South Africa and the Middle East, is
complicated there by the fact that there are four sets of players, each with its own moderate and hard-
line factions powerful enough to scuttle an agreement: Great Britain, the Irish Republic, the Ulster
Unionist Party, and the Social Democratic and Labour Party (SDLP) and Sinn Fein together on the
Nationalist side. Once the “external” players (Britain and Ireland) decided to work toward settlement,
the participation of the “internal” players was gradually achieved through a combination of political
sticks and carrots, guarantees, and pressure.
The 1985 Anglo-Irish Agreement marked the start of a new peace process by setting the stage
for a closer working relationship between Dublin and London. When Irish Taoiseach Garrett
FitzGerald came to power in 1982, he immediately began to shift relations with both Britain and the
North. The agreement between the governments required that the constitutional status of Northern
Ireland would not be changed without majority consent. It guaranteed the status of Northern Ireland as
part of the United Kingdom, but opened the door to the possibility that its status could be changed in
future. Significantly, the Agreement also “formalized cooperation on conflict resolution” between the
The Anglo Irish Agreement would eventually set the terms of the accord.
Albert Reynolds was elected Irish Taoiseach in 1992. He pushed “talks about talks” forward by
starting parallel dialogues with John Major’s government in Britain and with the SDLP and Sinn Fein
in Northern Ireland. In 1993 Major and Reynolds announced the Downing Street Declaration as the
starting point of a peace process. The British government reiterated that it had “no selfish strategic or
Yazid Sayigh, Armed Struggle and the Search for State, p. 660.
The difference between what Northern Ireland collects in revenue and what it receives in transfers from London is
the subvention. Northern Ireland’s is higher than Scotland or Wales’s. Interview with Dr. Esmond Birnie, former MLA,
UUP July 29, 2003.
Brendan O’Leary and John McGarry , The Politics of Antagonism: Understanding Northern Ireland (N.J.: Athlone:
London & Atlantic Heights(, 1996), p. 226.
John Darby, Scorpions in a Bottle (London: Minority Rights Group, 1997) and Eamonn Mallie and David
McCittrick, The Fight for Peace: The Secret Story Behind the Irish Peace Process (London: Heinemann, 1996).
economic interest in Northern Ireland,” and went on to acknowledge that the possibility of a united
Ireland was in the hands of the Irish people alone.
By signaling that it would not guarantee the
Unionist position indefinitely, Britain raised the cost of recalcitrance to Unionists. The Irish Republic
in turn promised that a settlement would include amending the Irish Constitution to remove the claim
that the Irish Parliament had in principle the right to incorporate and govern Northern Ireland. The
Declaration thus moved further toward establishing the framework within which both sides would
pursue their aspirations.
The Downing Street Declaration also stated that negotiations would be limited to those parties
not engaged in paramilitary violence. In response, the IRA announced a complete cessation of all
military activity in August 1994, forcing Loyalists to parry with a ceasefire. Within six months, the
British and Irish governments issued a Frameworks for the Future policy document, aimed at
translating the Joint Declaration into concrete terms. The guidelines for a final settlement included the
structure of relations between Ireland, Britain, and Northern Ireland, and the composition of a
devolved government within Northern Ireland. As in South Africa and the Middle East, then,
negotiations over a settlement in Northern Ireland were seriously underway by the early 1990s.
Understanding why the outcomes diverged as they did concerns us next.
3 Theory and practice of commitment
If negotiations are to lead to viable agreements, the adversaries must rely on one another. A
potential obstacle is the classic commitment problem described by Schelling: if each side knows that
the other might subsequently defect, why should either agree?
In theory, commitment problems are
ubiquitous in democratic politics, given the lack of third party enforcement. Despite numerous
attempts to show that compliance with democratic outcomes can be in the interests of all, no
theoretical account has been developed that shows why electoral losers with the power to defect so
often do not do so.
It would be unthinkable for an American president who lost an election to order
tanks down Pennsylvania Avenue, even if he has no realistic hope of ever regaining power through the
ballot box. The same could be said of politicians in many other democracies who routinely accept
John Darby, “The background to the peace process” (2003) http://cain.ulst.ac.uk/events/peace/darby03.htm
We are grateful to Brendan O’Leary for suggesting this interpretation of the effect of the Declaration on aspirations.
See his “Afterword: What is framed in the framework documents?” Ethnic and Racial Studies, Vol. 18 (1995), pp. 862-72.
Thomas C. Schelling, The Strategy of Conflict (Cambridge: Harvard University Press, 1960).
See Adam Przeworski, Democracy and the Market (Cambridge: Cambridge University Press, 1991) chapter one and
“Minimalist conception of democracy: A defense.” In Ian Shapiro and Casiano Hacker-Cordón, eds., Democracy’s Value
(Cambridge: Cambridge University Press, 1999), pp. 23-55.
results that consign them to political oblivion. We must therefore take care not to judge agreements in
the transition context by a standard that predicts perpetual civil war throughout the democratic world.
That said, there are reasons to expect commitment problems surrounding transplacements to be
particularly acute. Following decades of sometimes-violent conflict, they are marriages of convenience
among parties with little reason for mutual trust. As Rabin put it in 1993: “Peace is not made with
friends. Peace is made with enemies, some of whom—and I won’t name names—I loathe very
Even if reformers and moderates are willing to move toward agreement, they will be
skeptical of one another’s good faith.
This is further complicated in the quasi-democratic settings characteristic of SAMENI
transplacements because the negotiating partners must be responsive to public opinion. Indeed, it
might be possible for opponents of negotiations to use public opinion to undermine reformers or even
to remove them from power. Unless the reformers and moderates build support for the idea of an
agreement among the grass roots constituencies they depend on, the rug will be pulled out from under
them. Yet by the same token negotiators can also employ the constraints of democratic legitimation to
signal their commitment to a settlement. By making concessions public, political elites tie themselves
to positions they will not credibly be able to abandon without damaging their political careers. In so
doing they burn bridges to existing sources of legitimation, forcing them to look for new ones. In this
sense the quasi-democratic character of SAMENI negotiations may offer possibilities for dealing with
commitment problems that are not available in other transplacements.
Reformers and moderates have incentives to do what they can to help strengthen one another to
deal with hostile flanks, but these incentives are mixed. On the one hand they need to strengthen their
adversaries. Because negotiated transitions occur only when government reformers and opposition
moderates are too weak to achieve unilateral change but strong enough to achieve it if they cooperate,
they must have adversaries who can deliver. Yet, on the other hand, they must not strengthen their
adversaries too much. Both sides will want to extract the best possible terms for their supporters so far
as the content of an agreement is concerned, and they have no reason to make this task more difficult
than necessary. Moreover, in many cases the protagonists will expect to compete for political support
in the new order, if it arrives, and a stronger adversary is more difficult to compete with than a weaker
one. Even if the eventual settlement is expected to be a partition, other considerations create similar
pressures. Strengthening your adversary will turn out to have been costly if there is no agreement and
For elaboration see Shapiro, The State of Democratic Theory, pp. 88-93.
the situation reverts to one of open conflict. While each side has incentives to optimize its own
political strength vis-à-vis conservative and revolutionary flanks, it is thus optimal for them if their
adversary merely satisfices: becomes just strong enough to deliver an agreement from which potential
spoilers can be marginalized if they cannot be co-opted.
The capacity to demonstrate commitment is shaped by how bad things are likely to get should
the negotiations fail. If the principals believe that withdrawing from negotiations is either unlikely or
exceedingly costly for them, this will stiffen their backs to stay the course when the going gets rough.
More important, passing a costly or unacceptable reversion point helps them signal to their negotiating
adversaries that they are serious about achieving a successful agreement. This is why things sometimes
have to get worse before they can get better. An unpalatable reversion point for either or both parties
by no means guarantees agreement—there are other possibilities such as civil war or military coup.
But if your adversary knows that the status-quo ante is decreasingly tenable for you, it becomes easier
for him to believe that you are serious about looking for an accommodation.
Both sides must be concerned not only with an adversary’s political will to reach an
agreement, but also with their capacity to deliver. As a result, the credibility of negotiating
commitments is unavoidably dependent on how successful reformers and moderates are at co-opting or
marginalizing flank attacks. You have little reason to trust even an adversary you believe to be sincere
if you think that the ground may be cut from under him. This belief can be forestalled in various ways.
One is to actually be the flanking force. This Nixon-to-China logic suggests that the closer negotiators
are to the potential extremes in their parties, the more credible their commitments will be. The
alternative is to face down the flanking opposition at critical junctures, or visibly to burn bridges with
it while retaining the support of the military. One way or another, the negotiating principals must
ensure that their adversaries have good reasons to believe that they can deliver down the stretch.
3.1 Textbook success in South Africa
These commitment problems were managed in three ways in South Africa. First, the situation
on the ground became decreasingly attractive to the NP and eventually even to the ANC. The
combination of economic malaise, a sustained national uprising, and international opprobrium took an
increasing toll on white South Africans. The September 1992 Bisho massacre made graphic the
possibility that escalating violence could spiral out of control, forcing both sides to look into the abyss
“From Setbacks To Living Together,” The New York Times, 5 September 1993, p. 10.
See Jung and Shapiro, “South Africa’s negotiated transition,” pp. 280-282.
and resume (secret) negotiations that had been abandoned with the breakdown of CODESA II.
Second, potential flank attacks were effectively neutralized. De Klerk’s history as an orthodox
Afrikaner and conservative member of the NP initially strengthened his hand within his own party.
Similarly, Mandela had substantial political capital on entering the negotiations that stemmed from his
personal legitimacy. In contrast to Inkatha leader Mangosuthu Buthelezi, who was compromised by his
dealings with the apartheid regime in the 1980s, Mandela’s refusal to renounce violence as a condition
for release from 27 years in prison made him unassailable within the ANC. A radical flank did emerge
during negotiations, but Mandela’s legendary status and position as founder of the ANC’s military
wing made his authority impossible to challenge.
Third, and perhaps decisively, decisions made by Mandela and De Klerk early in the
negotiations helped diminish their commitment problems. De Klerk’s bold steps in 1992 showed how
bridge burning enhances credibility. In 1990 and 1991 he lost a series of by-elections while he was
negotiating with the ANC, emboldening right wing attacks on him. He called their bluff, however, by
holding a snap referendum in March 1992 that he won by a two-thirds majority among the white
electorate in every region of the country. He insisted throughout the referendum campaign that the
critical issue of power sharing (the political equivalent of sovereignty over Jerusalem for Israelis or
policing and IRA de-commissioning in Northern Ireland) was non-negotiable. In fact he was forced to
moderate this demand later in the negotiations, but even then this was obscured by the fact that the
ANC, which had steadfastly rejected all compromise on majority rule, gave De Klerk the wiggle room
he needed by agreeing to a constitutionally mandated interim government of national unity. The ANC
nonetheless refused to commit to a permanent government of national unity. They kept constitutionally
mandated power sharing out of the statement of entrenched principles the Constitutional Court would
eventually use to judge the acceptability of the final constitution. By the time the final constitution,
which dropped power sharing, was negotiated in 1995, the NP was no longer in a position to insist on
This was one of several respects in which the ANC played its cards perfectly during the
negotiations. At the time of the referendum, no one knew how long an interim settlement would last or
what the final agreement was going to look like. In many places interim settlements have been known
to drag on for decades, and some may reasonably have expected this in South Africa. Once De Klerk
had made his move, the ANC helped him satisfice by compromising on the power-sharing issue in the
Ibid., p. 288.
interim constitution. By then he had burned his bridges with the far right and legitimated the
negotiated transition among the white electorate, even though—fortuitously, as we argue later—it was
not entirely clear where it would lead.
Governments have an initial advantage in transition negotiations because they control the
military and hence the possibility of a return to authoritarianism should negotiations fail. However,
that advantage diminishes for a leader who alienates the conservative flank (which often has its own
links to the military hierarchy), and moves toward the position of his adversary during negotiations.
Facing down the hard right magnified De Klerk’s personal political investment in achieving the
successful negotiated settlement. Failure would have been immensely costly for him, possibly not
survivable. It would likely have been followed by a massive escalation of violence for which he would
have been held responsible by the whites who had trusted him, opening the way for an authoritarian
leader, or the army, to seize the initiative. We cannot be sure De Klerk had passed a point of no return
by the time of the referendum, but clearly he was well into his Rubicon treading in deep water, and
would have been in dire straits had he found himself there alone. The ANC pooh-poohed the
referendum at the time as one more illegitimate “whites only” vote, but once it was over they had De
Klerk exactly where they wanted him. He could no longer point to constraints coming from the right as
a way of limiting the concessions he could make. By the time of the final agreement in 1993 it was the
government that had made the decisive power-sharing concession, and by then there was no going
back for De Klerk.
The compromise over power sharing also illustrates how the opposition can deal successfully
with potentially hostile flanks within its own ranks. During negotiations, the initiative was not most
seriously in danger of shifting to more radical organizations outside the ANC (which by this point
were hopelessly weak), but rather to a radical flank of youth within the ANC mobilized by Winnie
Mandela, Peter Mokaba, and Chris Hani. Conceding power sharing for an interim settlement only
enabled the ANC leadership to keep critics on board at the critical meeting of February 1993.
moderate ANC leadership could plausibly (and correctly, as it turned out) make the case that time was
on their side, and that once the reality had changed on the ground, they would be negotiating over the
final settlement from a position of much greater strength.
To be sure, they were helped by a variety of factors: the ANC’s good organization compared
with that of potential opposition interlopers, Mandela’s legendary status, and the assassination of Chris
Ibid., pp. 290-91.
Hani in April 1993, which removed the most popular radical leader from the scene and reinforced the
commitment of both sides to a settlement. Hani’s murder might have unraveled the peace process
altogether as millions of African youth emptied into the streets to mourn and seek revenge. As it
happened however, De Klerk and Mandela moved quickly, and with a united front, to forestall such an
outcome. But the main reason the ANC coalition stayed together was that the moderate leadership,
which included tough-minded and pragmatic negotiators like Cyril Ramaphosa with unassailable anti-
apartheid credentials, could make a plausible case that by conceding power sharing in the interim
arrangement they had not conceded anything of importance. It meant that within four years the ANC
would have achieved an outcome through negotiations—majority rule democracy with themselves in
decisive control—that they lacked the military capacity to impose on the government at any time
before the transition.
If the ANC played its cards perfectly in the 1992 settlement, does this mean that they got the
better of De Klerk in the negotiations? Making that case would require establishing that the ANC
leadership could both have remained intact and continued as the principal protagonist on the
opposition side while agreeing to a permanent power-sharing arrangement. This is doubtful, not only
because of the internal conflict it would have provoked in the ANC, but also because any such deal
would have empowered Buthelezi’s Inkatha as an important opposition player with a share in power.
Inkatha was the third largest party in South Africa and the best bet for the NP to dilute ANC power
and support. The NP, which had long courted Inkatha as a moderate alternative to the ANC, would
have insisted on maximizing the strength of all minority parties with an eye to weakening the ANC.
Having successfully marginalized Buthelezi, the ANC had no reason to travel down that path.
“Ordinary democracy” rather than power sharing was thus their reservation price for the permanent
constitution, making their optimizing and satisficing strategies identical. Agreeing to interim power
sharing was needed to move De Klerk toward, if not past, his point of no return; resisting anything
more was essential to maintaining their own position. For this reason it seems clear that although De
Klerk could have scuttled the negotiations, paying whatever political price that entailed, he could not
have negotiated better terms for the NP.
Notice, however, that massive though the concession was to give away constitutionally
mandated power sharing in the final constitution, the outcome could have been worse for the NP. This
is not a negligible list of what they achieved: entrenched democratic principles with a constitutional
Inkatha played an important role in the government of national unity in the decade after the transition, but on
sufferance from the ANC which found it expedient to co-opt Buthelezi by keeping him in the cabinet.
court to interpret them; a two-thirds majority requirement to alter the constitution; entrenched
protections of property rights and civil freedoms; absence of high representation thresholds that would
disenfranchise minor parties; the guarantee of an amnesty process and protection of civil service jobs
for at least five years; and a powerful party whip system that the NP believed would strengthen its
leverage in Parliament. Two general elections later, ANC hegemony still means that there is little
meaningful national political competition.
If and when the ANC begins to fracture, leading to a more
fluid political environment, then the entrenched democratic guarantees will be important devices in
giving NP and its successor parties the chance to become more consequential political players.
Accordingly, it would be a mistake to say that De Klerk was giving away the store in accepting
constitutionally mandated power sharing. Arguably the NP made some unnecessary minor
concessions, and to that extent did not get the ANC to satisfice in areas where it might have done.
But on the major constitutional questions of democratic politics, commitments were extracted from the
ANC that might not have been, and indeed have not been in other transitional contexts.
This is not to say other outcomes are unimaginable. A more strategically astute NP leadership
in the 1980s might have made a deal with Buthelezi to partition the country before the UDF had
consolidated itself as the principal opposition player. Had that happened, the South African conflict
today might look more like the Middle East does, with disputes over sovereignty, borders, refugees
and other displaced persons and an endlessly debated “peace process” amid the ebb and flow of a war
of attrition. Alternatively, the NP could have staggered on during the 1990s (and perhaps even beyond)
as it had in the 1980s, repressing the opposition and thumbing its nose at outside pressure. This would
have meant an increasingly authoritarian militarized society with scant hope for improvements in the
economy, but it might have been survivable for a long time. True, the deteriorating political and
economic climate supplied the NP with the impetus to begin the negotiations and to stay the course,
but it did not compel them to do so. Structural factors predispose things in one direction or another, but
agency is required as well. Had PW Botha remained at the NP helm it is unlikely negotiations would
have started in 1990 or, if somehow they had begun, ended successfully in 1993.
3.2 Missed opportunity in the Middle East
However, there is meaningful local and regional political competition. In the Eastern Cape, the United Democratic
Movement (UDM) formed by Bantu Holomisa who was expelled from the ANC in 1997 and Roelf Meyer, who had led the
NP negotiating team in the transition, took away a substantial portion of the ANC vote to become the official opposition
party in 1999. The Christian Democratic Party did the same thing in the Northern Transvaal. In both 1994 and 1999 Inkatha
won elections in Natal, and the NP (in coalition with the DP after 1999) governs the Western Cape.
After the first election much of the NP defected to other opposition parties.
See, for instance, Jung and Shapiro, “South Africa’s negotiated transition,” pp. 300-301.
Though in some ways more challenging, the commitment problems facing the principals in the
Middle East in the early 1990s were not insuperable. Like Mandela and De Klerk in South Africa,
Rabin and Arafat were both well placed to manage hostile flanks. Israelis viewed Rabin as a war hero
dedicated to Israeli security, better positioning him to move the process forward than Shimon Peres
would be. Arafat also had the clout of being a long-time Fateh leader-in-exile. Indeed, Rabin’s
decision to deal with Arafat directly through the “back channels” at Oslo was a clear recognition that
the Palestinian delegation depended on and deferred to Arafat.
He risked a breakaway by his radical
flank, but he quickly demonstrated his ability to gain the acquiescence of the majority and to
marginalize the remaining opponents.
Yet the Middle East negotiations differed from those in South Africa from the perspectives of
both sides to the negotiations. The Oslo formulation was seen as Arafat’s attempt to shore up his
personal power, and it was far from clear that when push came to shove many Palestinians would
accept the agreement that could be extracted as superior to the status quo.
Nor was the Israeli side
propelled by an imperative to consummate an agreement. Israel’s decision to enter negotiations was
based, in part, on the perception that a weakened Arafat would be easy to bargain with. The intifada
and then significant US pressure had moved the Rabin government into concessions in 1993, with an
eye to getting an agreement rather than merely going through the interminable motions of the peace
process. Like Arafat, Rabin seems to have been personally committed to reaching a settlement, but few
can have doubted that Israel could cut and run if the going got rough.
This reality generated commitment problems that Arafat and Rabin both dealt with by burning
bridges on their flanks. Convinced that Rabin was serious about a settlement that would involve a
sovereign Palestinian state and focused, perhaps, on his own political survival, Arafat accepted a
partial agreement that postponed deciding the most important issues such as control over Jerusalem,
full establishment of a state, the return of refugees, and water rights.
He also stepped decisively into
the Rubicon by recognizing Israel’s right to exist and committing himself to policing the West Bank to
provide Israel security from Palestinian attacks. Opposition heightened, and the situation threatened to
turn into a civil war in November 1994 when Palestinian police faced down several thousand Hamas
supporters demonstrating outside the largest mosque in Gaza. In addition to the Islamist resistance,
Uri Savir, The Process, p. 5; David Makovsky, Making Peace with the PLO, p. 39.
Yazid Sayigh, Armed Struggle and the Search for a State, pp. 658-660.
According to Edward Said, as early as 1992 Arafat seemed to be “staking his entire future on Rabin’s electoral win.”
“Interview with Edward Said by Abdullah al-Sinnawi, Al-Arabi, Cairo (January 30, 1995), trans. Joseph Massad in Edward
Said, Peace and its Discontents (London: Random House, 1995), p. 179.
Arafat also faced increasing opposition from leftist and nationalist critics who decried the “Bantustan
Rabin also tried to face down hard-line opponents. Labour dominance in the Knesset allowed
him to move forward in negotiations with little real attention or recognition of opposition. But as he
converged on the peace settlement, signing Oslo II in September 1995, an increasingly vitriolic
opposition attacked his cooperation with Arafat, shouting “Death to Arafat” and portraying Rabin
wearing Arafat’s trademark kaffiyeh.
As Netanyahu remarked in that same month, “I don’t want to
say isolated, but we were in the minority. [Now] I think the government is in the minority.”
Rabin and Arafat demonstrated their commitment to the process, continuing to cooperate even
as they became targets of increasing attack. Rabin and Peres minimized Palestinian violations in an
attempt to shore up the process. After the April 9, 1995 bombings, for example, Arafat offered his
condolences while Rabin vowed to continue the peace process, telling Arafat, “We must work together
to prevent terrorism, and you must remember that terrorists are not just our enemies, but yours as
Arafat showed similar patience when on February 25, 1994, Dr. Baruch Goldstein from Qiryat
Arba walked into the Ibrahimi Mosque in Hebron’s Tomb of the Patriarchs and fired on worshipping
Muslims, killing 29 and wounding nearly 100. Rather than call off the process, Arafat recalled that he
told Rabin in a call following the massacre: “There are clearly fanatics in the settlements, and the
government of Israel needs to take steps against them. They want to destroy the peace process.”
reaction, too, was met with harsh Palestinian criticism.
Had they been able to keep going, Arafat and Rabin might well have consummated an
agreement whose benefits would have replenished their political capital on the South African model,
but Rabin’s murder in November 1995 put this possibility on ice. Peres’ failure to call a snap election
as soon as he replaced the fallen Rabin was surely one the most consequential missed opportunities in
the history of Middle East politics. It might well have served the same function as De Klerk’s March
1992 referendum, at a time when public opinion on both sides favored a two-state solution and the
Ibid.; Usher, Palestine in Crisis, pp. 14-20, 25-34.
On opposition in the Settlements, see Robert Friedman, “Report from the West Bank: An Unholy Rage,” The New
Yorker, March 7, 1994, pp. 54-56; Ehud Sprinzak, Brother vs. Brother; “Rabin decides to close Gazan roads near
settlements, Arafat Condemns Attacks,” New York Times, October 4, 1995; “Five killed in suicide bombing of bus 26 in
Jerusalem,” New York Times, August 21, 1995.
Barton Gellman, “Likud leader hammers Rabin, PLO Premier; Hopeful Netanyahu claims ascendancy of Israeli
opposition,” Washington Post, Sept. 9, 1995, p. A22.
“Rabin decides to close Gazan roads near settlements, Arafat Condemns Attacks;” see also Conni Bruck, “A reporter
at large: the wounds of peace” The New Yorker, October 8, 1995, pp. 64-91.
Janet Wallach and John Wallach, Arafat: In the Eyes of the Beholder (New York: Birch Lane, 1997), p. 460.
outrage at Rabin’s assassination had all but the most fanatical Israeli right on the defensive.
US pressure could have moved Peres at this point but the timing for this was not propitious in the run-
up to the November 1996 US presidential election.
True, many critical issues remained unresolved, but as the ambiguity around permanent power
sharing during the South African referendum underscores, this does not mean that a workable
settlement was unavailable. Indeed, in South Africa this ambiguity was essential to moving things
forward. Had the South African negotiations fallen apart, many analysts would subsequently have said
that the negotiations could not have succeeded because the whites would never have given up power
sharing. Plausible as such arguments might have sounded, they would have been wrong. Negotiations
themselves shift conceptions of what is possible, which in turn changes what is possible.
In the event, Peres missed his opportunity, tacking instead to the right. He permitted the
assassination of Yahya Ayyash in January 1996, which further contributed to the cycle of violence and
closures on the West Bank and Gaza Strip. In addition, Peres responded to attacks from Southern
Lebanon by bombing Lebanese refugee camps in Operation Grapes of Wrath. At the same time, the
wave of suicide bombings in the spring of 1996
led Israelis to seek a “firmer stance” in
Likud was able to regroup while Peres alienated himself from Israeli supporters of the
negotiations, particularly Israeli Palestinians. Palestinian radicals thus helped secure the victory of the
Israeli right, and when elections were held in May 1996, Peres lost the race to be Israel’s first directly
elected prime minister by only 29,000 votes. In marked contrast to Peres, the incoming Prime Minister
Binyamin Netanyahu had made no secret of his hostility to the Oslo accords on which the negotiations
had been predicated, leaving them on life-support at best.
Facing defeated partners in peace left Arafat massively weakened. Initially, he responded by
attempting to push the peace process forward at all costs. He courted Netanyahu, waiting for the call
and meeting that would eventually confer recognition on him as a legitimate negotiating partner.
According to polls conducted by the Tami Steinmetz Center for Peace Research (TSC) at the University of Tel Aviv,
the Oslo peace index of Israeli public opinion rose from 46.9 in October 1995 to 57.9 on November 8, immediately
following Rabin’s assassination. The index remained at 58 at the end of November, dropping to 55.8 in December. TSC,
“Peace Index, 1995,” http://spirit.tau.ac.il/socant/peace/ (6/18/03). 72.5 percent of Palestinians polled in October 1995
supported the peace process. JMCC Public Opinion Poll #10 http://www.jmcc.org/publicpoll/results/1995/no10.htm. For
additional discussion of the missed opportunities for settlements under Rabin and Peres, see Ellen Lust-Okar and Ken
Organski, “Coalitions and conflict: The case of Palestinian-Israeli negotiations over the west bank,” Journal of Conflict
Management and Peace Science (Spring, 2002).
By March 1996, Israelis had experienced 12 suicide bombings during the Oslo process. Four of these came in
February and March 1996 alone, killing 59 Israelis. Aish HaTorah, “Myths and Facts,”
Netanyahu, facing US pressure and an Israeli constituency pressing for a “secure peace,” finally
agreed not only to a meeting but also to signing the Hebron Accord and Wye Agreement. Arafat
responded with increased concessions, but it was clear that Netanyahu was neither a willing nor
committed partner in the peace process. Facing competing constituencies at home, he refused to
implement the agreement and continued expanding settlements in the Occupied Territories.
Palestinians became increasingly disillusioned with the peace process, and with the Palestinian
Authority’s (PA) ability to deliver a positive solution. The Palestinian standard of living had fallen
sharply since the signing of the 1993 Accords. Palestinians faced economic and social difficulties, and
an unresponsive, authoritarian PA. International support from donors dropped as they lost confidence
in the PA. Israel closed territories, continued expanding the settlements,
and dragged its feet in
withdrawing from the occupied territories.
This compounded Palestinian alienation.
Support for the
PA declined, with the majority of Palestinians coming to see it as corrupt.
The costs of selling any
agreement with compromises thus rose dramatically for Arafat after the collapse of the Hebron Accord
and Wye Agreement. Seventy percent of Palestinians polled in June 1999 continued to support the
peace process, but their trust in Israel had declined.
Arafat faced opposition not only from Hamas and
Islamic Jihad, but also from former Fateh supporters, academics, and the Palestinian middle classes
who had been willing at least to “wait and see” through 1996.
A June 1996 poll by the TSC found that 70.7 percent of Israeli Jews supported a firmer stance towards the
Palestinians. TSC, “Peace Index June 1996,” spirit.tau.ac.il/socant/peace/peaceindex/ 1996/files/JUNE96e.pdf (6/25/03).
The number of settlers rose from 99,065 in 1991 to 186,135 in 1999. Foundation for Middle East Peace, “Special
Report,” 11 (6), November-December 2001, http://www.fmep.org/reports/2001/v11n6.html (6/25/03).
Had the aborted Wye Memorandum of October 1998 been implemented, Palestinians would have had control over
18.2 percent (Area A) of the West Bank and shared control (Area B) over 21.8 percent. Israelis would maintain full control
(Area C) over 60 percent of the territory. Of the scheduled transfers, only 1 percent was territory moved from Area C to
Area A, and 3 percent of territory designated Area B would have remained in “nature reserves” in which Palestinians would
be prohibited from building. This agreement was signed after the initially declared deadline for the establishment of a
Palestinian state. Ministry of Foreign Affairs, “Wye Memorandum Agreement, October 23, 1998,”
Rex Brynen, A Very Political Economy: Peacebuilding and Foreign Aid in the West Bank and Gaza (Washington:
USIP, 2000), pp. 140-44.
A poll conducted by the Center for Palestine Research and Studies (CPRS) from 3-5 June 1999 found that 71 percent
of Palestinians believed the PA was corrupt, and 66 percent believed that the level of corruption would remain the same or
increase in the future. CPRS, “Public opinion poll # 41,” www.pcpsr.org/survey/cprspolls/99/poll41a.html (6/17/03).
The CPRS poll conducted from 3-5 June 1999 found that 70 percent of Palestinians surveyed supported the peace
process, while 27 percent opposed it. At the same time, however, 66 percent of the respondents did not trust the peaceful
intentions of the Barak government, in contrast to 23 percent expressing trust in the newly elected Israeli government.
Similarly, 55 percent did not believe that final status negotiations would lead successfully to a permanent settlement, and 45
percent supported the continuation of armed attacks against Israel. Ibid.
Thus, by the time Barak went to Camp David in 2000, intending to make major concessions,
Arafat could no longer meet him half way.
Elected in May 1999, Barak was riding a wave of anti-
Netanyahu sentiment and knew his landslide victory was largely in response to the Israeli demands to
“end this process”
and make an agreement.
As a directly elected prime minister, he enjoyed a degree
of independent legitimacy, and the possibility of holding a referendum on an agreement even in the
face of a hostile Knesset. That he was willing to make bold moves was readily apparent as, in an
attempt to force a peace agreement with the Syrians, he announced the unilateral withdrawal from
Southern Lebanon. The attempt to conclude an agreement with Hafez al-Asad was bold and creative. It
was made in the apparent belief that the Syrian treaty was easier to conclude than the Palestinian
agreement, and that peace with the Syrians would ease the way for the latter treaty.
After the strategy
failed, Barak turned fully to the Palestinian track in the spring of 2000, signaling his commitment to
make more concessions than any previous Israeli leader.
The resulting anger in the Knesset, marked
by a failed no-confidence vote and threats to bolt the coalition, bolstered his credibility.
The difficulty was that Arafat was now too weak to make concessions on such key issues as
Jerusalem and the Palestinian “right of return.” Palestinians’ skepticism toward Barak’s intentions had
only heightened since the previous summer, as they saw themselves sidelined in favor of Syrian-Israeli
negotiations. By the time he came to Camp David in the summer of 2000, against his will and under
strong pressure from Bill Clinton, his hands were tied. Survey data showed the majority of Palestinians
opposed to the meetings, with little confidence in Arafat’s negotiating team.
By then Arafat was
The extent to which concessions offered at Camp David were “major” and intended to meet Arafat half-way remains
controversial. However, it appears clear that these concessions went beyond previous Israeli offers (much to many Israelis’
dismay), and indeed exceeded offers which Arafat had previously found more acceptable.
Barak won the 1999 elections for prime minister with 56.08 percent of the popular vote, vs. 43.92 percent for
Netanyahu. See “Election Results 1999,” Jerusalem Post, http://info.jpost.com/1999/Supplements/Elections99/final.html
“Barak survives no-confidence vote as raids on Lebanon resume,” CNN.com World, February 14, 2000,
Thus William Safire would attack Barak for making concessions in violation of his own election pledges by offering
Arafat virtually all of the West Bank (including the Jordan Valley which would have meant relocating 40,000 Israeli
settlers), a virtual guarantee of a right of return to all Palestinians around the world, and shared sovereignty with a new
Palestinian state over portions of Jerusalem, “unthinkable only a year ago.” “Why is Arafat smiling?” The New York Times,
July 27, 2000, p. A25.
“Barak survives no-confidence vote in parliament,” CNN.com World, July 10, 2000.
A Jerusalem Media and Communications Centre (JMCC) poll conducted 16-17 July 2000 found that 52.8 percent of
respondents did not expect the delegations to reach an acceptable final agreement, while only 37.3 percent expected an
agreement. More importantly, when asked “Are you confident or not confident in the Palestinian negotiating delegation in
Camp David?” 34.7 percent lacked confidence and 7.8 percent “did not know.” JMCC, “JMCC public opinion poll no. 38
on Palestinian attitudes towards the Camp David Summit, July 2000,” www.jmcc.org/publicpoll/results/2000/no38.html
likely unable to restore the levels of support he had obtained in the early 1990s; his only hope of
maintaining Fateh dominance and his leadership position was by responding to Palestinian popular
Unless Barak was willing to concede to Palestinian demands, which seemed vanishingly
unlikely, Arafat would thus be unable to respond. This was clear to the lead writers for The Economist
a week before the negotiations collapsed. Citing opinion polls giving only 32 percent support among
Palestinians (with over 50 percent believing that he would be pressured into concessions at Camp
David), they noted with great perspicacity that “the more he withstands the heat, the higher his stock
Former Secretary of State James Baker reached the same conclusion in his post-mortem
following the collapse. Quoting Palestinian sources to the effect that “Arafat’s ability to maneuver is
he concluded: “what was not enough for Mr. Arafat was too much for many Israelis, to whom
any agreement will be submitted by referendum.”
In short, the window of opportunity was no longer
open because Arafat was not in a position to commit to anything that Barak could accept.
The Israeli-Palestinian negotiations demonstrate how easily potentially viable solutions to the
commitment problem can be destroyed. Rabin had put himself and his leadership on the line, first in
making and then defending his decision to negotiate with the PLO. His solid, if slim, Labour majority
in the Knesset enabled him to act with little regard for his opponents. The bridges he burned along the
way suggested that he would do what was needed to deliver an agreement once made. Arafat, who had
his own reasons to negotiate, could thus anticipate that if he took the risk and signed the Oslo
Agreement the Israeli government would fulfill its part of the bargain and move forward on the final
status issues. Rabin took similar risks, even if somewhat less was at stake for him initially. He had
reason to believe that Arafat would deliver on his commitments, given the political costs he had paid
for entering negotiations. Certainly it was clear that if anyone could deliver the Palestinian side in
1995, it was he.
But the derailing of the process eroded his political power, and with it his ability to
deliver the Palestinian side in any agreement. Opponents of Oslo grew from small Islamic and leftist
According to a Palestinian Center for Policy and Survey Research (PCPSR) poll, 68 percent of Palestinians believed
Arafat’s overall position at Camp David was “just right,” while 15 percent believed he had compromised too much. PCPSR,
“Public opinion poll #1,” 27-29 July 2000, www.pcpsr.org/survey/polls/2000/pla.html (6/17/03).
“The ballad of Camp David,” The Economist, July 2, 2000, p.43.
For data supporting this conclusion, see PCPSR, “Public opinion poll # 1.”
James A. Baker III, “Peace, one step at a time,” The New York Times, July 27, 2000, p. A25.
Even those who question whether Arafat turned Hamas and the Islamic Jihad “loose” on the Israelis argue that he
did so in order to increase his bargaining position and ultimately gain a better settlement. Few seriously question whether he
really sought to end the conflict and see a Palestinian state develop before his imminent death. As he made clear in his
March 2002 interview with Christiane Amanpour, he sees the establishment of the Palestinian state (including East
Jerusalem) not only as his personal mission, but as the very definition of who he is. See
fringe groups into the mainstream of Palestinians, who came to believe that years of interim
agreements weakened them while providing no benefits. They continued to support “peace,” but by
2000 the vast majority did not expect the then current process to succeed.
3.3 Northern Irish vulnerability to multiple vetoes
Negotiations in Northern Ireland have faced two limitations. Moderate leaders have not had the
success of Mandela and De Klerk in facing down their radical flanks, particularly on the Unionist side,
and neither side has been sufficiently motivated by a deteriorating status quo to take the irreversible
steps to consolidate agreement. The absence of an intifada or other serious threat to governability
makes it remarkable that there has been an agreement at all, but unsurprising that the agreement has
been in perpetual danger of falling apart.
Frustrated by the failure of the British government and the Unionists to negotiate seriously, the
IRA suspended its ceasefire in February 1996 with a bomb explosion that injured 100 people in
As in South Africa, violence was effective in galvanizing a recalcitrant negotiating partner,
and three weeks later the Irish and British governments announced that inclusive all party negotiations
on Northern Ireland would follow elections to a negotiating Forum. Chaired by former US Senator
George Mitchell, talks began in June 1996 under rules of “sufficient consensus,” so that no proposal
could pass if vetoed by Britain, Ireland, the Unionist UUP or the Nationalist SDLP.
These talks remained bogged down in party brinksmanship and infighting until a Labour
landslide put Tony Blair into office in June 1997. Blair was free of the ties and debts to Unionist
parliamentary partners that had hampered John Major’s room to move the peace process. Blair
immediately expressed his commitment to “solving” the Northern Ireland crisis. His first trip was to
Northern Ireland, where he warned Sinn Fein that “the settlement train is leaving. I want you on that
train. But it is leaving anyway and I will not allow it to wait for you. You cannot hold the process to
ransom any longer. So end the violence now.”
In June and July the British government worked hard
behind the scenes to bring Sinn Fein into talks, to the growing ire of Unionists who feared Britain
In July 2000, 75 percent of Palestinians polled supported the Palestinian-Israeli peace process, but 60 percent
believed that lasting peace was impossible. Moreover, 66 percent of Palestinians polled also believed that Israelis do not
believe lasting peace is possible with Palestinians. PCPSR, “Public opinion poll # 1.”
Richard W. Stevenson “Bomb Wounds 100 in London as IRA Truce is Said to End,” The New York Times 2/10/96
The idea of sufficient consensus was drawn directly from the South African CODESA negotiations, and predictably
facilitated Unionist stonewalling. See Robert H. Mnookin “Strategic Barriers to Dispute Resolution: A Comparison of
Bilateral and Multilateral Negotiations”Journal of the Institute of Theoretical Economics, Vol. 159, No. 1 (2003).
George J. Mitchell, Making Peace (New York: Alfred Knopf, 1999), p. 101.
would sell them out to achieve peace. But the British and Irish governments recognized that no
settlement would be enforceable without Sinn Fein’s participation. Here the negotiators made a key
decision that Israel and the US had not made in 2003 when they sought to marginalize Hamas from the
road map: to include all potential spoilers to the agreement.
But bringing Sinn Fein into talks jeopardized the political strength and negotiating position of
Trimble’s UUP. If Unionist support for negotiations crumbled, forcing Trimble to leave the table, the
peace process would disintegrate. Moderators and guarantors were therefore careful to shore up the
Unionist side, to protect it from its own right wing.
Moderators catered to the Unionist demand for
IRA arms decommissioning, for example, by giving the issue prominence at the start of the talks.
The IRA responded by announcing a second ceasefire on July 20, 1997, while continuing to
refuse to decommission. Since the Unionists had made decommissioning a precondition of
negotiations, Trimble took the risky decision to enter talks that could have gutted his support base. He
had evidently reached a personal point of no return, as demonstrated by his private admission to Blair
that “we are not in the mode of walking out.”
On July 22, the debate over whether Sinn Fein should
be admitted to talks without prior decommissioning came to a head in a vote. The UUP, DUP, and
UKUP all voted against it, with the result that when talks reconvened in September Sinn Fein was at
the table but the Unionist parties were gone.
The DUP and UKUP had left for good, and tried to
force the UUP to walk out through accusations that the party was betraying its people. Opinion polls
showed that the UUP had popular support for remaining in the talks, but the party leadership was also
under extreme pressure, even from within its own ranks.
This was the situation when the UUP finally
entered negotiations under Trimble’s leadership. Talks between the governments and the parties began
seriously in October 1997.
Negotiations were organized in three strands. The first dealt with political arrangements
within Northern Ireland, the second with North-South relations, and the third with relations between
London and Dublin. Strand Two, concerning the relation between Ireland and Northern Ireland, was
the most contested. In February, all parties agreed to an Easter deadline, and after a delay caused by
ceasefire violations on both sides, the parties began serious negotiations in mid-March. As they came
No agreement was in fact reached over decommissioning however, which of course continued to act as a stumbling
block to implementation as late as 2003.
Mitchell, Making Peace., p. 108.
Ibid., p. 109.
Ibid., pp.111, 117.
down to the wire, London and Dublin negotiated an agreement on Strand Two that was blatantly
unacceptable to the Unionist side. At the insistence of the moderators, who argued that Trimble was
not bluffing when he said he could not agree to this document, both sides returned to the table to
Strand Two was reworked in the final week before the deadline to include the
controversial provisions demanded by each side: a North-South council and an elected Assembly
expected to operate in mutual interdependence. The Good Friday Agreement, also known as the
Belfast Agreement, was concluded in April 1998.
The accord included five main constitutional principles: Northern Ireland’s future
constitutional status, as part of Ireland or the U.K., would be in the hands of its citizens; The people of
Ireland, north and south, could vote to unite; Northern Ireland would remain, for the time being, within
the United Kingdom; Northern Irish citizens could choose to identify as Irish, British, or both; And the
Irish state would drop its territorial claim on Northern Ireland.
A copy of the Agreement was
delivered to every household in Northern Ireland in anticipation of the referendum, and a clear
majority in both Ireland (56 percent turnout; 94 percent approval) and Northern Ireland (81 percent
turnout; 71 percent approval) approved it. This included majorities of both Unionists and Nationalists,
though the Unionist majority was slim.
As with the Oslo Accord, but unlike the ANC-NP agreement, the Belfast Agreement did not
mark the end of negotiations and the beginning of implementation. For almost two years, London
continued to govern Northern Ireland as implementation snagged on the controversial issues that had
been left outstanding in the Good Friday agreement. It seems clear that part of the obstacle to
implementing the Belfast Agreement was that a substantial portion of Unionists never believed that the
status quo was unsustainable. Unlike in South Africa, where polls indicated that most whites had
concluded apartheid was no longer feasible by the end of the 1980s, most Protestants in Northern
Ireland continue to believe direct rule from London, or majority rule in which they would be the
majority, are sustainable alternatives.
A compromise with Nationalists could only weaken their
John Darby, “Northern Ireland: The background to the Peace Process,” CAIN Web Service
http://cain.ulst.ac.uk/events/peace/darby03.htm (June 5, 2003), p. 7.
Paul Bew “Initiative to Trimble but his edge over opponents is thin,” The Path to Peace, website, April, 1998
http://www.ireland.om/special/peace/results/analysis/analysis10.htm (5/25/2000) p. 1. 55% of Protestants approved the
In late 2000 The Northern Ireland Life and Times survey found that 14% of Protestants who voted “yes” in 1998
would now vote against the Agreement because of declining support for devolution and lack of progress on
The failure of implementation is partly explained by the fact that Trimble and the moderate
Unionists have been unable to face down the right wing that opposes agreement. From the outset,
moderate Unionists had only a narrow margin of support for the accord. Exit polls from the May 1998
referendum on the agreement showed Protestants almost evenly divided between support and
opposition. The Protestant middle class appeared ready to defect from the settlement over the early
release of prisoners.
Whereas the March 1992 South African referendum returned a solid
endorsement for continued negotiations, the results of the Northern Ireland referendum were
sufficiently ambiguous that they could still be used to political advantage by those who opposed a
settlement, highlighting again the fickle role of democracy in negotiations.
Although 96 percent of
Catholics supported the agreement, only 55 percent of Protestants did.
Moreover, the Protestant vote
was exceedingly fragile. According to the Northern Ireland Referendum and Election Study, one
quarter of all Protestants had considered changing their vote during the campaign, mostly from a Yes
to a No vote. Among Catholics, only seven percent had considered changing theirs.
Elections for the Northern Ireland Assembly were held in June 1998. The UUP won 28 seats,
the SDLP 24, DUP 20, SF 18, Alliance 6, Women’s Coalition 2, UKUP 5, PUP 2, and Anti-
Agreement Unionists 3. Although the UUP won a plurality of seats (not of votes), its slim margin of
victory actually represented a loss for the party, whose 1997 returns at Westminster presaged a win of
as many as 39 seats, and only 16 for Ian Paisley’s DUP. The 1998 election results suggested that the
moderate center of Unionism was eroding. Moreover, the UUP was comparatively vulnerable as the
anti-agreement camp (DUP, UKUP, and AAU) also won, between them, 28 seats. Analysts predicted
before the election that the UUP would need to win at least 30 seats to avoid deadlock in the Assembly
and to make the North/South council work.
Therefore, although the UUP won the election and
decommissioning. Hayes and McAllister, “Who voted for Peace?” and Bernadette Hayes and Lizanne Dowds
“Underpinning Opinions: Declining Levels of support Among Protestants for the Good Friday Agreement,” Paper presented
at roundtable discussion by Democratic Dialogue 10 April 2001, Europa Hotel, Belfast
Ibid., p.2; Frank Millar, “London is relieved but difficulties lie ahead”
http://www.ireland.om/special/peace/results/road/ahead3.htm p.1 (5/25/2000).
Suzanne Breen, “United No parties set their sights on Assembly,”
Sydney Elliot and W.D. Flackes. Conflict in Northern Ireland: An Encyclopedia, (Belfast: Blackstaff Press, 1999, p.
Bernadette C. Hayes and Ian McAllister, “Who Voted for Peace? Public Support for the 1998 Northern Ireland
Agreement, Irish Political Studies, Vol.16 2001 (PSAI Press), p. 81.
Gerry Moriarty, “How the parties could share out seats,” Path to Peace website,
http://www.ireland.om/special/peace/results/road/ahead4.htm (5/25/2000), p. 2.
emerged as the largest party in the Assembly, anti-agreement parties were also able to interpret the
election result as a victory, in particular because the transfer system of voting favored the UUP.
The election campaign also laid bare differences within the UUP over the agreement. Jeffrey
Donaldson, a UUP Member of Parliament at Westminster, emerged as the most important opposition
figure within the party but almost half of the leadership of the UUP openly opposed the accord. Some
of these took seats in the Assembly, but they could not be counted on to vote the party line, further
diluting the pro-agreement bloc.
Trimble barely squeaked by in elections for party leadership after
1998, as he faced powerful challenges over the issue of implementation of the accord. When Trimble
agreed in 2000 that it might be possible to re-enter government with Sinn Fein without prior
decommissioning, Martin Smyth mounted an internal challenge for party leadership which, though
unsuccessful, exposed deep and continuing fault-lines within the Ulster Unionist Party and the
weakness of Trimble’s support base.
Under pressure from Britain, Trimble nevertheless twice
entered a power-sharing government without IRA decommissioning.
On August 15, 1998 a bomb exploded in Omagh, killing 28 people (mostly women and
children) and injuring 220—the largest loss of life of any single act of violence during the Troubles.
The Real IRA, a breakaway faction of the IRA opposed to the settlement, claimed responsibility and
immediately apologized for the deaths, announcing the suspension of all military operations. But the
act was so widely condemned, including by the IRA, that it backfired against hard-line Republicans,
reminding everyone of the gruesome alternative to peace. Omagh played an important role in
solidifying a commitment to peace across Northern Ireland and in permanently marginalizing those
from the Nationalist right who would scuttle the process. Since then, Sinn Fein and the IRA have
faced less of the hard-line censure that has limited Trimble’s room to maneuver in the implementation
phase of the process.
The UUP blocked Sinn Fein’s entry to the Executive for 16 months, insisting again on prior
IRA decommissioning, while Britain, Ireland, and mediators continued to try to broker a
In November 1999, a slim majority of 58 percent of UUP delegates approved entry into
Gerry Moriarty, “Even split for anti, pro-agreement parties,”
Paul Bew, “Initiative to Trimble but his edge over opponents is thin,” The Path to Peace, website, April, 1998
http://www.ireland.om/special/peace/results/analysis/analysis10.htm (5/25/2000), p. 1.
Jonathan Tonge and Jocelyn A.J. Evans, “Faultlines in Unionism: Division and Dissent Within the Ulster Unionist
Council,” Irish Political Studies, Vol.16 (2001), pp.113-114.
At the time, Trimble insisted that his hands were tied, and Ken Maginnis, a leading moderate within the party,
agreed that the plan would split the party and was impossible to sell to rank and file supporters. Shawn Pogtchnik,
a joint government, and the British and Irish governments transferred power from London to Belfast
within days. Under a power-sharing formula, Trimble became First Minister, and Seamus Mallon,
leader of the Nationalist SDLP, the co-equal Deputy First Minister. Ten other cabinet seats were
divided proportionally among Ulster Unionists, the SDLP, Sinn Fein and the DUP. The UUP was left
deeply divided by the split vote, and the terms of entry included a clause committing the party council
to reconvene in February to review the decision. If the IRA had not by then begun to disarm, the party
would use its majority position to dismantle the government.
The IRA did not move on decommissioning, however, and London suspended the Assembly in
February 2000 to protect Trimble from another divisive UUP vote that threatened to sink his
leadership. Hours after the suspension, the IRA made its first commitment to dismantle its arsenal.
March, Trimble won an internal challenge to his party leadership with only 57 percent support,
indicating sustained divisions over power sharing and further compromising his room to maneuver.
In a bid to face down his opponents, Trimble announced his intention to sever ties with the Loyalist
Orange Order. Although the move was part of a longstanding agenda to reduce the party’s sectarian
connections, the timing was clearly calculated to hive off naysayers within the party.
After the IRA agreed to weapons dump inspections by international assessors Cyril
Ramaphosa and Martti Ahtisaari, and to put their guns and bombs beyond use, 66 percent of Unionists
polled in May 2000 said that they wanted Trimble to return to a power-sharing government with Sinn
The UUC finally voted at the end of May to resume implementation, and Britain transferred
power back to the Northern Ireland Assembly and its 12-person joint Cabinet.
Devolution was suspended three more times over the next three years, so that the power-
sharing government in Northern Ireland was operational for only twenty of the possible fifty-four
“Ultimatum Irks Northern Ireland” The Detroit News, Sunday, July 4, 1999
Warren Hoge, “Ulster Unionists Open Way for Ruling with Sinn Fein,” The New York Times 11/28/99
Warren Hoge, “Britain Suspends System of Power Sharing in Ulster,” The New York Times 2/12/2000
Warren Hoge, “Ulster Leader Holds on, but Power Lessens in Unionist Vote,” The New York Times 3/26/2000
Frank Millar, “Trimble stakes his future on the drive to modernise,” The Irish Times, 4/21/2000
Suzanne Breen, “Most Ulster Unionists want Trimble in Executive,” The Irish Times, May 12, 2000
Reuters “Northern Ireland Rivals Try for Home Rule Again” The New York Times 5/30/2000
months it might have functioned.
In September 2002 the Ulster Unionist Council issued a statement
affirming that the UUP would “not sit in government with unreconstructed terrorists,” again
demanding the total disbandment of all terrorist groups including the IRA.
In October 2002
devolution was suspended again because Trimble threatened to withdraw his Ministers from the
Executive in protest against the continued participation of Sinn Fein in government.
repeatedly backed the Unionist position, noting that the Irish Republic has refused to let Sinn Fein take
seats in government without IRA dissolution, whereas Unionists in the North had been forced into a
power-sharing arrangement with Sinn Fein.
Gerry Adams complained that London should not have
the latitude to suspend the institutions of the Good Friday Agreement every time the UUP threatened
to walk out, pointing out that Blair would be unlikely to do the same if Sinn Fein threatened to leave
Trimble made the contrary case that Britain should not have suspended devolved
government; that it ought instead to have suspended Sinn Fein.
Post-agreement negotiations have nevertheless settled a number of other issues that were left
outstanding in April 1998. In November 2001 the Royal Ulster Constabulary was re-named the Police
Service of Northern Ireland, and the new Policing Board replaced the Police Authority. Although the
UUP contested the arrangement, the PSNI committed to recruiting on the basis of 50:50
This is potentially important. In South Africa the integration of the police force
before the transition greatly eased enforcement problems. The IRA also made an important symbolic
statement in July 2002, apologizing for killing and injuring non-combatants during the Troubles.
IRA decommissioning persisted as the main obstacle to implementation however, leading analysts to
assume that the relevant question was “what would have induced (or might in future induce) the IRA
to disband as a paramilitary organization?”
Text of a Speech made by Gerry Adams, Monaghan, 26/10/2002, CAIN Web Service
http://cain.ulst.ac.uk/events/peace/docs/ga261002.htm (11/11/2002) p. 6.
Text of Document agreed at the Ulster Unionist Council meeting on Saturday 21 September, 2002; CAIN Web
Service http://cain.ulst.ac.uk/issues/politics/docs/uup/uup210902.htm (11/11/2002).
This occurred after a scandal in which Sinn Fein was accused of spying on other parties and government at
Speech by Tony Blair, Harbor Commissioners’ Office, Belfast, 10/17/02 CAIN Web Service
Speech made by Gerry Adams, Monaghan 10/26/02, CAIN Web Service
CAIN Web Service, Chronology of the Conflict, http://cain.ulst.ac.uk/othelem/chron/ch00.htm (06/05/03).
Full text of IRA statement of apology, 16 July 2002; CAIN Web Service
But the South African experience suggests otherwise. South Africa’s NP also insisted initially
on disarmament as a pre-condition for talks with the ANC, but when the ANC refused, talks proceeded
almost without a pause. Considering that an international monitoring team repeatedly verified that IRA
weapons were beyond use
and that Loyalists perpetrated most of the post-accord violence,
it is far
from obvious that the IRA’s refusal to change its rhetoric was a literal threat to security in Northern
Ireland or Britain in the late 1990s.
Yet Tony Blair has repeatedly suspended power-sharing
governments out of sensitivity to Unionist demands. For instance, in May 2003 he again halted
implementation by postponing legislative elections. He deemed insufficient Gerry Adams’ statement
that the IRA would “disarm fully as part of the Northern Ireland peace settlement if other parties to the
accord fulfill their obligations,”
despite declarations by the Irish government and most pro-
Agreement parties that the insistence on disarmament caused more problems than it solved. American
envoy Richard Haass also called for elections to take place “as soon as possible.”
By July 2003,
many politicians in Northern Ireland, including pro-Agreement Unionists, agreed that postponing
elections placed tremendous pressure on the settlement and seriously undermined the credibility of the
agreement among voters.
Pro-agreement members of the Ulster Unionist Party were particularly
concerned that party in-fighting would mean that postponed elections could favor those who opposed
SAMENI negotiations are prone to forms of myopia that lead pressure to be applied in the
wrong places. Blair’s repeated suspension of power sharing was a failure of vision and nerve that
ranks with Peres’ missed opportunity in 1995, playing into the hands of those Unionists for whom
stonewalling against change has always been the name of the game. This is not to say that great and
continuing pressure on the IRA would not have been needed to move the process forward. It is to say
that Blair was uniquely placed, among recent British prime ministers, to bring no less essential
pressure to bear on the Unionists. They too had to be told that the “settlement train is leaving the
The IRA suspended contact with the Independent International Commission on Decommissioning (IICD) on 30
October, 2002, full text of IRA statement, CAIN Web Service http://cain.ulst.ac.uk/events/peace/docs/ira301002.htm
Hayes and McAllister, “Who Voted for Peace,” p. 88.
DUP party member David Ervine argues for example that it was destructive of the UUP to insist on a condition that
they should have known the IRA would be unable to comply with. (Interview, June 20, 2003).
Warren Hoge “Sinn Fein leader Pledges Full Disarmament of the I.R.A.” New York Times (4/28/2003), p.A2.
Brian Lavery “US Envoy Wants Elections” New York Times, (5/8/2003), p. A2.
Interviews with David McNarry and Dr. Esmond Birnie (UUP), Rachel Steert (Women’s Coalition), David Ervine
(DUP), and James Cooper (UUP).
Interviews with Dr. Esmond Birnie and James Cooper (UUP).
station.” The relevant South African precedent here is not the failed CODESA idea of “sufficient
consensus” which empowers and emboldens spoilers, and which, notably, fell apart twice without
moving the process forward. Rather it is that powerful players committed to a settlement must make
spoilers realize that they will be marginalized if they do not join the process. Recall that the NP and
ANC negotiated the core elements of the South African settlement in secret in 1992, and then
announced them as non-negotiable. Inkatha agreed to add its name to the ballot only days before the
1994 election, once when Buthelezi finally realized that it could not be derailed.
This merits particular note in light of the fact that in other areas there have been moves toward
the type of normal politics in Northern Ireland that are the ultimate goal of SAMENI negotiations.
When government has been in session, much (though not all) of what goes on is politics as usual.
Statements and proposed legislation about road safety, healthcare, unemployment, and pork barrel-type
projects dominate the websites of all the major parties. The Women’s Coalition and the Alliance Party
explicitly eschew sectarian designation and are self-consciously attempting to generate a new political
dialogue that will expand the possibilities of political identity in Northern Ireland. Polls indicate that a
majority of Protestants and Catholics alike had also moved beyond the conflict in the new century,
agreeing that issues like health service and unemployment were most pressing.
To sum up, the obstacles to peace in Northern Ireland were as formidable in the 1990s as they
were in the Middle East. Yet our analysis suggests that had different choices been made at critical
junctures, things could have turned out differently, and there might have been settlements comparable
to what was achieved in South Africa. Indeed, underscoring the fluidity of SAMENI negotiations and
their critical dependence on contingencies of choice, we saw that the South African transition might
well have derailed at various points, just as the others have done.
4 From what could have been to what could be
There is nothing intrinsic to the conflicts in the Middle East and Northern Ireland that renders
them less tractable than South Africa’s. People miss this either because they focus on the wrong
features of the conflicts, or because they focus on the wrong features of negotiations. Recognizing
SAMENI conflicts as a particular type of transplacement, in which democratic legitimation is relied on
to achieve agreement and make it stick, allows us to elaborate on the conditions that make settlements
more and less likely, and to discuss the implications for the Middle East and Northern Ireland.
In the 1999-2000 Northern Ireland Life and Times survey, respondents were asked to identify the most important
priorities for the new assembly. Forty percent chose improving health services and 37 percent cited employment.
4.1 SAMENI negotiations revisited
Perhaps the most powerful challenges to our reasoning would come from the Middle East.
Some will contend that the conflict there is fundamentally different from that in Northern Ireland or
South Africa on the grounds that the stakes for the actors—both Israelis seeking to defend a Jewish
state and Palestinians invested in returning to their homeland—are inherently zero sum. The great
lesson of the wars of religion of the seventeenth century is often taken to be that when national
sovereignty becomes bound up with collective religious identities, the result is perpetual war. The
standard solution has long been thought to lie in the de-emphasis of such exclusionary grounds for
citizenship via mechanisms such as religious disestablishment. Most players in, and commentators
about, the Middle East operate on the assumption that this is not possible there. It is said to be too
threatening to the Zionist self-understanding, and to its mirror image: that the Palestinian people have
an inalienable right to national self-determination.
But consider a South African perspective on this objection. Apartheid was self-consciously
exclusionary, built on an ascriptive basis that left no room for conversion. Moreover, the racial
ideology of apartheid was underpinned with a religious mission; its architects were doctrinaire
Calvinists who saw themselves as one of the last outposts of Christian civilization—defending it from
communism in the east and a corrupt and degenerate west. As recently as 1985, had anyone seriously
suggested that white South Africans would endorse a multiracial state—let alone under a majority-rule,
black government—they would have been laughed out of town or locked up. Nor should we forget
that, whereas five million Jews face a similar number of Palestinians in Israel, Gaza, and the West
Bank, in South Africa five million whites faced twenty-five million blacks. Demography alone might
lead one to believe that the zero-sum character of the South African conflict in the mid-1980s looked
less tractable than is the case in the Middle East. The magnitude of the unexpected South African
transformation suggests that analogous changes in beliefs about religion and ethnicity might indeed be
possible in the Middle East.
A different lack-of-comparability objection focuses less on the stakes involved in the conflict
and more on the proposed solution. One reason for this claim is that the two-state solution will never
In this connection it is perhaps heartening that at a conference on democratic transitions and consolidation
consisting of some 100 academic experts from 36 countries plus 33 heads and former heads of state held in Madrid in
October/November 2001, a final report was adopted in which it was agreed that “rights of citizenship should apply equally
to all citizens” and that the majority “must avoid all temptation to define the nation in ethnic terms in the constitutional text
or its political practice.” See Diego Hidalgo, ed., Conference on Democratic Transitions and Consolidation (Madrid, Spain:
Siddharth Mehta Ediciones, 2002) p. 34. See also Ari Shavit, “Cry, the beloved two state solution,” Haaretz, August 7, 2003.
be perceived as legitimate. Manifest disparities of wealth, status, and power, combined with the partial
character of most proposed variants of the Palestinian state, call its sustainability into question. It is
doubtless true that any two-state solution would confront legitimation problems reminiscent of
“separate but equal” in the American south. But this does not mean that it could not endure for a long
time; indeed, the American example suggests that it could. It simply suggests that in the longer term it
may be a weigh station en route to a different destination. Living with a settlement can change what
people can live with. Brown v. Board of Education would not have been possible in 1896.
Another objection to the two-state solution is that it is unenforceable. Any future Israeli
government would remain free to roll tanks into the West Bank or Gaza if it became unhappy with the
settlement. In contrast, this reasoning goes, in South Africa the enforcement problem was dealt with by
a fundamental shift in power to the ANC. Recall, however, our caution at the outset of §3 that the
enforcement difficulties associated with SAMENI negotiations should not be judged by standards that
are not met in most political circumstances. Switzerland and Costa Rica are substantially demilitarized
states that could be invaded by neighbors but are not. Moreover, it is not obvious that the enforcement
problem was solved by the ANC’s triumph in South Africa. The army might have defected at various
points—as indeed it still could.
At issue is less whether there is a two-state solution and more, whatever the solution, that it
gains enough legitimacy that potential spoilers decide that challenging it is too costly. It is far from
clear that this could not have occurred had the negotiations between Rabin and Arafat been able to
conclude in 1995, or had Peres adopted a different policy than he did immediately following Rabin’s
death. We saw in §3.2 that at that time the two-state solution enjoyed considerable legitimacy. Blair
was in an even stronger position to solve the enforcement problem in Northern Ireland after 1998,
given his historic mandate in 1997 and the support for the agreement in both communities. In the end,
it is only the legitimacy of the agreement itself that can get potential spoilers to adhere to the
conditions of a peace settlement.
Because they often underestimate the importance of democratic legitimacy in any negotiated
settlement, analysts often focus on, and pin their hopes upon, moderate leaders who are willing to push
a peace agreement forward but lack the standing to face down radicals. For instance, Gerry Adams’s
emergence on the scene in 1988 as a new kind of pragmatic IRA leader “with a human face” led to
endless speculation about whether a settlement could now be anticipated. But Adams and his behavior
were irrelevant until the Blair government came to power in the UK in 1997—given the dependence of
the Tory governments on Unionist support under Thatcher and Major.
Likewise, both US and Israeli negotiators have often focused on Palestinian players who could
not be expected to deliver. The 2003 Middle East road map was a case in point. Arafat’s corrupt
government had lacked grassroots legitimacy since 1995, and the appointment of Mahmoud Abbas
was scarcely a solution to this problem. With its political legitimacy, not to mention its security
apparatus, in tatters, the PA was in no position to rein in the violence that followed the Aqaba
The more Abbas was praised as “reasonable” in Jerusalem and Washington, the weaker he
was bound to become in Ramallah. Caught between a rock and a hard place, any popularity he could
hope to sustain would depend on delivering what Arafat could not: better living conditions, an end to
curfews, and open borders—all of which depend on Israeli good will and US pressure.
The presence of a strong radical flank need not itself bode poorly for peace, provided efforts
are made to incorporate them if they cannot be marginalized. That Hamas was not at the table in the
early negotiations over the road map is the functional equivalent of the South African government’s
trying to strike a deal with Buthelezi in 1992, or of talks in Northern Ireland excluding Sinn Fein.
Notwithstanding IRA failure to decommission as late as 2003, Sinn Fein support for the agreement
was unwavering, as its electoral base grew in the years following the accord. Hamas and other violent
Palestinian nationalist groups are doubtless aware that their strategy will never lead to outright victory
over Israel, but their immediate target is not Israel. They are engaged in a struggle for control over the
The Oslo process weakened Fateh and the PA vis-à-vis Hamas, which both provided critical social services to an
increasingly impoverished Palestinian people. [See Ian Fisher, “Defining Hamas: Roots in charity and branches of
violence,” New York Times, June 16, 2003.] Support for Hamas grew at Fateh’s expense. In December 1996, support for
Fateh was 35.2 percent and for Hamas 10.3 percent. Similarly, 41.2 percent of Palestinians most trusted Yasser Arafat, and
4.8 percent trusted Sheikh Yassin. Only 19.5 percent of Palestinians did not trust anyone. By December 2001, support for
Fateh dropped to 26.1 percent, while that for Hamas rose to 21.3 percent. Similarly, trust in Arafat declined to 24.5 percent
and that in Yassin rose to 12.8 percent. This trend has continued. By April 2003, Fateh remained the single most-trusted
faction in Palestinian politics, with 22.6 percent, although overall support for Fateh trailed the combined support for Hamas
and leftist factions (22.0 percent), Islamic Jihad (6.3 percent), PFLP (2.0 percent), and other factions (3.1 percent). However,
more 34.3 percent responded they “don’t trust anyone.” [Ibid.; JMCC, “Public Opinion Poll #43,”
http://www.jmcc.org/publicpoll/results/2001/no43.htm (6/25/03); JMCC, “Public Opinion Poll #18,”
http://www.jmcc.org/publicpoll/results/1996/no18.htm (6/25/03).] A poll conducted by JMCC in April 2003 found that the
majority of Palestinians believed that Abbas’s appointment as Prime Minister would have little effect on the PA reform
process: of respondents, 28.7 percent felt the appointment would further PA reforms, 17.4 percent felt it would hinder
reforms, and 43.2 percent felt it would have no effect. It is telling that 67.8 percent of respondents believed that the creation
of a prime ministry was due to external influences alone, 19.2 percent felt that it was due to external influences and a
conviction that the reforms were in the interests of the Palestinian people, and only 6.2 percent believed the reform was
undertaken purely in order to benefit the Palestinian people. Furthermore, only 1.8 percent of respondents named Abu
Mazen as the Palestinian personality that they most trust, vs. 21.1 percent for Yasser Arafat and 9.7 percent for Ahmed
Yassin. [JMCC, “JMCC Public Opinion Poll # 48,” April 2003,
representation of the Palestinians, and failure of the peace process has solidified its support base.
Effective marginalization of Hamas was likely impossible by 2003, so that any lasting peace must
involve dealing with them, perhaps in secret—at least initially.
Contrary to press reports at the time,
Sharon’s overture to the Palestinian Authority was not a case of Nixon going to China.
It was more
like Nixon being dragged to Hong Kong.
Even when the right actors are on board, they may have to sell mutually incompatible solutions
to their constituencies—at least until people realize that they can live with outcomes they had
previously dismissed as unthinkable. Constructive ambiguity can help. There seems to be little
question that in the Middle East obscurity about the final outcome, and even the interim steps, was
essential for moving the process forward in 1995. Lack of clarity about such issues as the final status
of Jerusalem, borders, settlements, and the right of return has been harshly criticized, but ambiguity
about these issues was essential to creating a new reality in which Palestinians and Israelis accepted a
two-state solution as legitimate. The South African success depended on the final agreement to
abandon constitutionally mandated power sharing not being fully apparent earlier in negotiations.
As the South African process illustrates, the very fact of participating in negotiations can
loosen up fixed perceptions. This is not to deny that constructive ambiguity can create implementation
problems later. Northern Ireland and the Middle East have revealed all too clearly that it can. But
without it they will not move forward, and the creative ingenuity of the players will never become
focused on dealing with the implementation problems.
If the SAMENI cases suggest that negotiations proceed best if the final details are left for later,
they also suggest that there is urgency in getting to an agreement. That the South African negotiations
moved quickly and decisively contributed greatly to the result. Conversely, the slowing down of
Middle East negotiations has repeatedly strengthened the hands of stonewallers and led windows of
opportunity to close. The Bush administration seemed to appreciate this in the summer of 2003, when
Secretary of State Colin Powell argued forcefully for the need to “move urgently,” not giving time for
By April 2003, Fateh remained the single most-trusted faction in Palestinian politics, with 22.6 percent, although
overall support for Fateh trailed the combined support for Hamas (22.0 percent), Islamic Jihad (6.3 percent), PFLP (.0
percent), and other factions (3.1 percent). However, more 34.3 percent responded they “don’t trust anyone.” Ibid.
Hamas moderates have signaled their willingness to enter negotiations. See Amira Hass, “What the doctor orders,”
Haaretz, June 20, 2003. http://www.haaretz.com/hasen/pages/ShArt.jhtml?itemNo=307203&. (6/22/03).
See John Diamond, “Sharon’s support for road map historic,” USA Today, May 26, 2003, p.1. See
the “terrorists to win.”
Even the most committed moderates must outrun the radicals and
reactionaries who will be determined to prevent an agreement. The sooner a negotiated arrangement is
seen as the new status quo, the less likely it is that recalcitrant forces will be able to destroy it.
4.2 Future prospects
Taken together, these observations suggest important lessons about the future prospects for
negotiations in the Middle East and Northern Ireland. Despite continuing setbacks to the
implementation of the Good Friday agreement, the Northern Ireland conflict appears considerably
closer to resolution than the Middle East. Once decommissioning is seen in perspective, it becomes
clear that the window of opportunity has been open for a good part of the time after 1998. Members of
the Legislative Assembly speculate that support for the agreement has eroded in part precisely because
it has been suspended so often that it seems unworkable.
The fact that it is suspended by
Westminster makes it seem additionally undemocratic in the sense of being more vulnerable to
external than electoral pressure. Blair was unwilling through 2004 to pay the political cost of putting
real pressure on the Unionists, but he had both the mandate and the leeway so to do. Moreover, it
seems reasonable to think that he might have done so, and that there may well be future opportunities
for him or others to implement the Good Friday agreement—particularly for a Labour government
with a large majority.
In 2004 there seemed to be three possibilities for Northern Ireland. Implementation could
continue to move forward in fits and starts, but with everyday politics revolving less around the
agreement and more around tangible improvements in areas like healthcare and employment. Polls
show that popular support for the Agreement has been diminishing ever since it was signed, but also
that most people in Northern Ireland believe that social welfare and the economy are more pressing
issues than the relative political status of Unionists and Republicans.
It remains to be seen whether
the voting public will be able to move the parties toward a similar consensus. Alternatively, if the
stalemate on decommissioning persists, there might be another big international push (possibly again
led by the US) to reach agreement on the outstanding issues. Publicity, attention, and deadlines would
be used again, as they were in 1998, to generate a groundswell of support and excitement for
“Powell condemns Hamas role,” June 21, 2003, 00:53 GMT 01:53 UK, BBC News,
Author interview with Esmond Birnie. A poll conducted in January 2003 showed that only 36% of Protestants
would still vote for the Good Friday Agreement. But 60% of Protestants would be willing to support the peace process if it
could be made to work. Colin Irwin, “Devolution and the State of the Northern Ireland Peace Process,”
See footnote 113 http://www.qub.ac.uk/ss/csr/nilt (06/05/2003)
implementation. Another possibility is a reversion to violence—even if the trend toward normal
politics against the background of the Good Friday agreement makes this outcome less likely. But
completely ruling it out assumes more backbone from the British government than Blair exhibited in
facing down Unionist recalcitrance in his first five years in office, and it takes too static a view of the
IRA leadership which could always revert to a military strategy.
Successfully maneuvering through the Middle East road map would require a great deal from
Palestinian, Israeli and international actors. The toll of the intifada, as well as the changing regional
situation after the US-Iraq war, led several key players to return to negotiations. At the same time,
however, these same conditions limited the trust each side can hold in the other, and hence the
likelihood that moderates can deliver an agreement. Prime Minister Sharon held his position largely
due to the weakness of the traditional Israeli peace camp. Given this reality, not to mention his history
and ideology, he was unlikely to become a reformer.
Indeed, his 2002 decision to construct a
massive fence in the West Bank was less likely meant to ensure the end of terrorism—at which it
failed—than to make conditions unbearable, thus stimulating Palestinian emigration from the occupied
Yet, if he appeared less than eager to negotiate, he was capable of delivering.
Deteriorating conditions had led many settlers to announce that they would accept evacuation from the
settlements in return for economic compensation, thus weakening the right flank.
If Sharon chose to
enter into the negotiations full-steam ahead, he could sideline his right wing supporters and offer
Labor participation in his government.
As the historical champion of the peace process they would
have little choice but to join. Sharon would have burned his bridges behind him, enhancing, from the
standpoint of our analyses, the prospects for peace.
We should not be surprised, therefore, by his equivocal statements on Israel’s “occupation” of the West Bank in
May and June 2003. John Kifner, “The Bush Plan: Put the Toughest Hurdles First,” The New York Times 6/8/2003 Section
4, pp. 1, 14; David K. Shipler, “Sharon Has a Map. Can He Redraw It?” The New York Times 6/1/2003 Section 4, p. 1.
Edward R.F.Shaheen, “The map and the fence,” The New York Review of Books, 50 (11), July 3, 2003,
http://www.nybooks.com/articles/16411 (6/25/03). The fence is expected to disrupt the lives of 600,000 West Bank
Palestinians. See “U.N. Estimates Israeli Barrier Will Disrupt Lives of 600,000,” New York Times, November 12, 2003.
According to a 2003 poll, 54 percent of settlers stated that they would resist forced dismantlement, but 74 percent
would entertain moving inside the Green Line in return for compensation. Moreover, 71 percent of settlers thought a peace
agreement should be reached (up from about 55 percent in 2002) and 44 percent accepted a Palestinian state (up from 19
percent the previous year). Israelis also increasingly see settlers as isolated in their struggle to secure settlement. The same
survey found that about 64 percent of settlers expect themselves to be alone, and 75 percent of Israelis inside the Green Line
see settlers as isolated in their struggle with the government. “Poll: 71% of settlers say there will be deal with the
Palestinians” Ha’aretz, July 23, 2003.
Ellen Lust-Okar, “Israeli elections 2003: what Likud’s victory doesn’t tell you,” Yale Israel Journal, Spring 2003,
Success would require significant US pressure on the Israeli administration to make
concessions in the face of continuing right-wing domestic opposition and Palestinian violence. The
Israeli government would eventually need to convince its own supporters that they are better off with
an agreement with the Palestinians and withdrawal from the occupied territories than they are with
either continued occupation of the West Bank and Gaza Strip (WBGS) or the ethnic cleansing of Eretz
Israel. This had become a tall order by 2002, when Israeli support for the “transfer” of Palestinians
living in the WBGS reached 46 percent, and the support for transfer of Palestinians living inside the
Green Line stood at 31 percent.
Perhaps even more difficult, however, was that Israelis also needed
to convince Palestinians that they would no longer solve immediate crises by rolling tanks into
Palestinian towns or assassinating leaders with military aircraft. Without that conviction, it is difficult
to see how support for any settlement would be forthcoming.
The tasks facing the Palestinians were equally difficult. Those intent on a settlement would
need to restrain the radical flank, demonstrating to Israelis not only that they could, but that they
consistently would, ensure Israeli security. This was particularly difficult at a time when the security
infrastructure of the Palestinian Authority had largely been demolished, and when support for the
Fateh leadership and confidence in the Israelis were both at record lows. With a strong radical flank,
the PA could not afford to sideline Hamas completely. Rather, it would need to gain at least tacit
Hamas support, most likely by granting Hamas moderates what they have most fervently wanted: a
place at the table.
Because even this would be unlikely to satisfy the hardliners who wanted only to
see Israel pushed into the sea, the need for security remained. A successful solution would thus require
a significant change in the attitudes of both Palestinian and Israeli hardliners in the long run.
The difficulty of the task should not surprise us. We have seen that the windows of opportunity
that make settlements possible open rarely, and they seldom stay open for long. Few politicians are
willing to take the considerable risks involved in moving through them. Indeed, they often fail to see
either the possibilities or how fleeting they might be. A better and more widespread understanding of
“Memorandum No. 61: Israeli public opinion on national security,” Jaffee Center for Strategic Studies, Tel Aviv
University, July 2002. See also “Growing Popularity Of A Transfer In Israel, Removing Palestinians From The West Bank
And Gaza And Distributing Them Throughout The Arab World,” National Public Radio Morning Edition, October 21, 2002.
Hamas moderates have signaled their willingness to enter negotiations. Amira Hass, “What the doctor orders,”
Haaretz, June 20, 2003, http://www.haaretz.com/hasen/pages/ShArt.jhtml?itemNo=307203& (6/22/03)
People-to-people programs and efforts to revise history textbooks on both sides are part of the efforts to change
fundamental attitudes toward “the other” on both sides. See “The Politics of Palestinian Textbooks” in 31, no. 1 (Autumn
2001): 5-19; Israel/Palestine Center for Research and Information program on Peace Education (see
the dynamics of SAMENI negotiations, and of their consistency with the logic of transplacements,
might diminish that possibility.
5. Concluding comments
Conflict studies have been driven in part by a debate, often implicit, over whether conflict is
driven from above—by political elites manipulating followers to gross acts of violence—or from
below—by ancient and primordial hatreds nurtured in families, communities, and places of worship.
The study of conflict resolution is similarly riveted, and focuses either on elite dispositions to negotiate
or on grassroots initiatives to foster tolerance. One of the functions of the imperfect democratic
settings we study in these cases is to draw the link, both empirically and analytically, between the two
levels of focus.
In the middle of a peace process, a society lacks both the security of a hard line and retaliation
(which represent the status quo ante), and the legitimacy of a democratic settlement (which is the final
goal). Suspended thus between an unsustainable past and an unreachable future, elite politics becomes
polarized. When the politics of the conflict—a fairly clear delineation between two sides—gives way
to the politics of peace—in which each side breaks into two or more factions—the most important
contests are those that take place among the factions. And in these fights, the primary weapon is
In both the Middle East and Northern Ireland the way forward seems primarily constrained not
by elites or masses alone, but by the link between them. In both places those factions that reject a
settlement, or that reject the particular settlement that is on the table, have gained ascendance since the
heady days (1994 in the Middle East, 1998 in Northern Ireland) when a majority on all sides favored
reconciliation. As negotiations have dragged on, a reforming center has lost ground partly because its
members have failed to lock in their advantage by making peace work.
Facing failure, or at most limited success, people grow impatient, or unnerved. Skittish, they
swing among factions. In the last four Israeli elections voters returned Rabin, Netanyahu, Barak, and
Sharon—in that order. David Trimble has been operating on an exceedingly thin margin of support
almost continuously since 1998, the Nationalist SDLP has lost support to Sinn Fein, and Arafat and the
PLO have grown steadily more vulnerable to Hamas and Islamic Jihad since the late 1980s. Electoral
volatility and slim margins of support stop any side from declaring a mandate in such circumstances,
http://www.ipcri.org/index1.html and Jerusalem Post (Internet edition), “Teachers greet the enemy,” April 3, 2003 at
undermining the chances of any significant move in one direction or another. Then democracy, the link
between elites and a support base, can paralyze transformation rather than facilitate it.
But paralysis is no more sustainable than the status quo ante was once considered to be, and
the specter of failure will continue to hang over these societies so long as some democratic peace is not
reached. The factionalized character of the politics of peace draws the link between leaders and
constituents even more starkly than normal politics, and their mutual dependence offers the possibility
of both vicious and virtuous cycles. A decisive swing in favor of a settlement will require some
success—some sense that moving forward is better than moving backward. Success lies in the hands
of elites. But support for an agreement, or for implementation, which is what is needed before elites
can move decisively toward peace without risk of losing power altogether, lies with their
constituencies. It goes without saying that some decisive break with the vicious cycles that currently
characterize these peace processes is a precondition for progress.