Two decades after the fall of Soviet-bloc dictatorships, popular movements for democracy are erupting in the last regional bastion of authoritarianism: the Arab world.

So far, only Tunisia's dictator, Zine el-Abidine Ben Ali, has been toppled, while Egypt's President Hosni Mubarak - who has ruled that ancient land longer than many pharaohs - announced Tuesday that he will step down in September. But other Arab autocrats are bound to go. From Algeria to Syria to Jordan, people are fed up with stagnation and injustice, and are mobilizing for democratic change.

So, what happens when the autocrat is gone? Will the end of despotism give way to chaos - as happened when Mobutu Sese Seko was toppled in 1997 after more than 30 years in power in Zaire? Will the military or some civilian strongman fill the void with a new autocracy - as occurred after the overthrow of Arab monarchs in Egypt and Iraq in the 1950s, and as has been the norm in most of the world until recently? Or can some of the Arab nations produce real democracy - as we saw in most of Eastern Europe and about half the states of sub-Saharan Africa? Regime transitions are uncertain affairs. But since the mid-1970s, more than 60 countries have found their way to democracy. Some have done so in circumstances of rapid upheaval that offer lessons for reformers in Tunisia, Egypt and other Arab countries today.

Unite the democratic opposition.

When a dictatorship is on the ropes, one thing that can rescue it is a divided opposition. That is why autocrats so frequently foster those divisions, secretly funding a proliferation of opposition parties. Even extremely corrupt rulers may generate significant electoral support - not the thumping majorities they claim, but enough to steal an election - when the opposition is splintered.

In the Philippines in 1986, Nicaragua in 1990 and Ukraine in 2004, the opposition united around the candidacies of Corazon Aquino, Violeta Chamorro and Viktor Yushchenko, respectively. Broad fronts such as these - as well as the Concertacion movement that swept Christian Democrat Patricio Aylwin to power in Chile in 1989 after the departure of Gen. Augusto Pinochet - often span deep personal and ideological differences. But the time for democratic forces to debate those matters is later, once the old order is defeated and democratic institutions have been established.

Egypt is fortunate - it has an obvious alternative leader, Mohamed ElBaradei, whom disparate opposition elements seem to be rallying around. Whether the next presidential election is held on schedule in September or moved up, ElBaradei, or anyone like him leading a broad opposition front, will probably win a resounding victory over anyone connected to Mubarak's National Democratic Party.

Make sure the old order really is gone.

The exit of a long-ruling strongman, such as Ben Ali, does not necessarily mean the end of a regime. Fallen dictators often leave behind robust political and security machines. No autocrat in modern times met a more immediate fate than Romania's Nicolae Ceausescu, who was executed by a firing squad of his own soldiers in 1989 just three days after a popular revolution forced him to flee the capital. Yet his successor, Ion Iliescu, was a corrupt former communist who obstructed political reform. Most of the former Soviet states, such as Georgia and Kazakhstan, had similar experiences.

Countries are much more likely to get to democracy quickly if they identify and embrace political leaders who are untainted by the old order and are ready to roll it back.

But also come to an understanding with the old order.

Victorious democrats won't be able to completely excise the pillars of the authoritarian order. Instead, for their country to turn toward democracy, those pillars must be neutralized or co-opted. This old order may descend into violence when, as in Iraq, broad classes of elites are stigmatized and ousted from their positions. In a successful bargain, most old-regime elites retain their freedom, assets and often their jobs but accept the new rules of the democratic game.

Unless the military collapses in defeat, as it did in Greece in 1974 and in Argentina after the Falklands War, it must be persuaded to at least tolerate a new democratic order. In the short run, that means guaranteeing the military significant autonomy, as well as immunity from prosecution for its crimes. Over time, civilian democratic control of the military can be extended incrementally, as was done masterfully in Brazil in the 1980s and in Chile during the 1990s. But if the professional military feels threatened and demeaned from the start, the transition is in trouble.

The same principle applies to surviving elements of the state security apparatus, the bureaucracy and the ruling party. In South Africa, for example, old-regime elements received amnesty for their human rights abuses in exchange for fully disclosing what they had done. In this and other successful transitions, top officials were replaced, but most state bureaucrats kept their jobs.

Rewrite the rules.

A new democratic government needs a new constitution, but it can't be drawn up too hastily. Meanwhile, some key provisions can be altered expeditiously, either by legislation, interim executive fiat or national consensus.

In Spain, the path to democratization was opened by the Law for Political Reform, adopted by the parliament within a year of dictator Francisco Franco's death in 1975. Poland adopted a package of amendments in 1992, only after it had elected a new parliament and a new president, Lech Walesa; a new constitution followed in 1997. South Africa enacted an interim constitution to govern the country while it undertook an ambitious constitution-writing process with wide popular consultation - which is the ideal arrangement.

An urgent priority, though, is to rewrite the rules so that free and fair elections are possible. This must happen before democratic elections can be held in Egypt and Tunisia. In transitions toward democracy, there is a strong case for including as many political players as possible. This requires some form of proportional representation to ensure that emerging small parties can have a stake in the new order, while minimizing the organizational advantage of the former ruling party. In the 2005 elections in Iraq, proportional representation ensured a seat at the table for smaller minority and liberal parties that could never have won a plurality in individual districts.

Isolate the extremes.

That said, not everyone can or should be brought into the new democratic order. Prosecuting particularly venal members of a former ruling family, such as those tied to the Philippines' Ferdinand Marcos, Indonesia's fallen strongman Suharto or now Tunisia's Ben Ali, can be part of a larger reconciliation strategy. But the circle of punishment must be drawn narrowly. It may even help the transition to drive a wedge between a few old-regime cronies and the bulk of the establishment, many of whom may harbor grievances against "the family."

A transitional government should aim for inclusion, and should test the democratic commitment of dubious players rather than inadvertently induce them to become violent opponents. However, groups that refuse to renounce violence as a means of obtaining power, or that reject the legitimacy of democracy, have no place in the new order. That provision was part of the wisdom of the postwar German constitution.

Transitions are full of opportunists, charlatans and erstwhile autocrats who enter the new political field with no commitment to democracy. Every democratic transition that has endured - from Spain and Portugal to Chile, South Africa and now hopefully Indonesia - has tread this path.

Fragile democracies become stable when people who once had no use for democracy embrace it as the only game in town.

The toppling of a brutal, corrupt, and long-ruling dictator, Zine el Abidine ben Ali, is an extraordinary achievement for the diverse elements of Tunisian society who came out into the streets in recent weeks to demand change. Ben Ali's startling fall is another reminder of how suddenly political change can come in authoritarian regimes that substitute force, fear, and fraud for legitimacy. Such regimes may appear stable for very long periods of time, but when the people lose their fear and the army refuses to fire on the people, they can unravel very quickly.

Unfortunately, the demise of a dictator does not guarantee the rise of a democracy in its place.  Historically, most authoritarian regimes have given way to a new (and often only slightly reconstituted) autocracy. This has been the principle pattern not only in the successor states to the Soviet Union, but in much of Africa since independence, and in numerous states in Asia and Latin America historically as well. In the Middle East, the odds against a successful democratic transition are particularly long, since there have hardly been any (outside Turkey and Israel) since the end of colonial rule. In Iran in 1979, a popular uprising against a long-serving dictator led not to democracy but rather to an even more odious and murderous form of oppression.

 If Tunisia is to defy the odds, it will need a significant period of time to reform the corrupt rules and institutions of the authoritarian regime and create an open, pluralistic society and party system that is capable of structuring democratic competition. Even if elections for a successor government are pushed out to six months, rather than sixty days, it is highly unlikely that this will provide sufficient time to create even a minimally fair and functional democratic playing field. 

Think of the many components of a democratic election, and Tunisia today is far from having them in place. After decades of fixed and phony elections, Tunisia needs a complete overhaul of its electoral machinery: a new and impartial electoral administration, a new electoral register, and perhaps as well a new electoral system. An energetic program of civic education should help Tunisians understand not only the mechanics of a democratic electoral process but also the underlying norms, rights, and responsibilities of democratic citizenship. This is a long process, but from Poland to Chile to South Africa, civil society organizations have shown that much can be accomplished to lay the foundations for popular democratic awareness and capacity if the models, materials, and resources are made available, and if there is a decent interval of time and political space to do the work. Doing this work-and enabling political parties and candidates to convey their messages-also requires a new and more pluralistic media environment. State control of the electronic and print media must be radically refashioned.  Privately owned media must be allowed to form and function, and critics of the old order must be allowed to enter the arena of ownership.

An effective democratic election requires not just freedom of opposition parties to organize, but time, resources, and training for them to form-or reform-and develop some ability to perform the essential functions of modern parties:  to establish what they stand for, to develop programmatic agendas, to elect leaders and recruit candidates, to forge ties with constituencies, and to survey public opinion and respond with appropriate messages. Trade unions, business chambers, and other civic groups need time as well to purge themselves of corrupting ties to the old order, or form anew, cultivate their natural constituencies, and build an authentic civil society.  Independent think tanks and public opinion surveys can also help to structure and enrich an emergent democratic process, but they as well need time and resources to function effectively.

Free and fair elections-especially in a context where they have never taken place before-also require extensive preparations for domestic monitoring and international observation, so that fraud can be detected and deterred, honest mistakes can be exposed and corrected, and public confidence can be generated in the new procedures.

Many of these tasks are ongoing after a successful transition to democracy, and setting too ambitious an agenda for reform could risk waiting indefinitely and squandering the opportunity for democratic change. But one of the most common reasons for failed transitions is a rush to early national elections and a failure to prepare the ground adequately for a fair and meaningful contest. Two common consequences of hurried elections are chaos or renewed autocracy, as some portions of the old order rally behind a new figure or old party and win by hook or crook. 

Unfortunately, there are also risks in waiting too long. Democratic energy in society can dissipate.  If (putative) democratic forces enter into a broad-based transitional government, as is now happening in Tunisia, they risk being corrupted or tainted with the stench of the old order if they hang around for too long, sharing some authority and stature but no real power. A prolonged transitional period can also give authoritarian forces time to regroup, purge the worst elements, present cosmetic changes, divide and confuse the opposition, and return to power under the guise of a pseudo-democracy. That is why it is important that opposition figures in Tunisia insist on a serious program of institutional and possibly constitutional reform during the transitional period, with extensive public dialogue and broad popular participation, so that interim rule is not a stagnant pause but rather a dynamic historical moment that engages and mobilizes public opinion for real democratic change. The risks of delay could also be reduced if a non-partisan, technocratic figure, not associated with the Ben Ali's political machine, could be tapped to lead the interim government, and if the political opposition could unify to negotiate strong conditions for the period of interim rule, including basic freedoms, an end to censorship, and removal of Ben Ali loyalists from the cabinet.

There is an important role for international actors at this seminal moment in Tunisian history. Like peoples throughout the Middle East and other post-colonial spaces, Tunisians are understandably wary of foreign intervention. After a quarter-century of lavish Western (especially French) aid and political comfort to Ben Ali, Tunisians will no doubt cast a suspicious eye on grants, statements and actions that purport to now, suddenly, want to build democracy in Tunisia. But Tunisians may welcome limited and specific steps if they are transparent and taken in careful consultation with diverse elements of Tunisia's civil society and historic opposition. 

Fortunately, Tunisia has many liberal and democratic figures in business, intellectual, cultural, and civic life who understand what liberal democracy is and would like to see it emerge in Tunisia. And it has other distinct advantages. It is a relatively small country in size and population, which makes some of the tasks of institution building and promotion of democratic norms a bit easier. Educational levels are relatively high, and there is a significant infrastructure of a middle class society. The security forces seem to be divided, and it appears the army refused to fire on peaceful protestors-a very positive precedent. Without blood on its hands from the recent violence, the army is better poised than other elements of state security to guarantee a process of democratic change, if its leadership comes down in favor of it (for whatever reason). And in contrast to Algeria, Egypt, or Jordan, Islamists do not seem to have strong public support. Thus, it is difficult for the forces of the ancien regime to manipulate public fears of radical Islam (or of disorder that the old elites themselves covertly generate) in order to discredit liberalism as naïve and ride back to power. 

It is vital that Europe and the United States not fall again for the specter of disorder or an Islamist surge, but rather insist on genuine democratic reforms, and tie future aid and geopolitical support to this. The US and EU should hold forth the prospect of Tunisia achieving a special and potentially transformative status in economic relations if it negotiates the path to become the first Arab democracy of this era. At the same time, they should threaten to institute targeted travel and financial sanctions against diehard defenders of the old order who frustrate or sabotage a democratic transition, or who use violence against peaceful demonstrators.  These kinds of prospective inducements, positive and negative, can help to tip the balance in the calculations of a lot of elites from outside the Ben Ali "family" but who were part of the Ben Ali regime and must now be wondering where their own interests lie. To complement the necessary private messages, the US ambassador (and others representing democracies in Tunis) should stand up publicly for democratic reforms, embrace democratic reformers, support new democratic initiatives with small grants, and warn old regime elites against repression.

In the coming weeks and months, American and European democracy foundations and aid organizations, along with the United Nations and its political assistance programs in the UNDP, can do a lot-transparently, and in consultation with Tunisian society-to train and support the emerging infrastructure of democracy in the state administration, political parties, and civil society. The funding required to make a difference is not large in absolute terms, and it should be a priority. Time is of the essence, and more flexible instruments, like USAID's Office of Transition Initiatives, should be tapped to activate assistance quickly.        

History-and the grim realities of pervasive authoritarianism in what is known in the political science discipline as a "bad neighborhood"-do not justify a high degree of optimism about the prospects for democracy in Tunisia. Yet the third wave of global democratization is replete with instances of successful democratization in even more unlikely circumstances. The speed with which the Tunisian protests mushroomed in a few weeks from a lone act of self-sacrifice to a national uprising, and the intensity with which this uprising has resonated in nearby countries, shows the pent-up demand for democratic change in the Arab world. If that demand can be directed toward pursuit of concrete institutional reform, with timely international support, the Jasmine Revolution could surprise again, by giving birth to the first Arab democracy of our time.