Thus far, most of the endless talk about the war in Iraq has focused on several issues: the scale of the operation, Washington's motivation, and the rift in the Atlantic alliance. It is now safe to assume, however, that if and when war comes (as of this writing, the battle had yet to begin), the United States and its allies will win, Saddam Hussein and his cronies will be toppled, and some sort of massive military occupation will follow.

In the aftermath of the war, the occupiers will focus on immediate tasks, such as ensuring order, providing relief to the long-suffering Iraqi people, and asserting control over the country. Very quickly, however -- even before they have met these goals -- the victorious powers will have to answer another pressing question: How, exactly, should they go about rebuilding the country? Saying simply that postwar Iraq should be democratic will be the easy part. Just about everyone agrees on that, and indeed, for many this end will justify the entire operation. The more difficult question will be how to make it happen.

Fortunately, the job of building democracy in Iraq, although difficult, may not be quite as hard as many critics of the war have warned. Iraq today possesses several features that will facilitate the reconstruction effort. Despite Saddam's long repression, democratic institutions are not entirely alien to the country. Under the Hashemite monarchy, which ruled from 1921 until 1958, Iraq adopted a parliamentary system modeled on that of its colonial master, the United Kingdom. Political parties existed, even in the opposition, and dissent and disagreement were generally tolerated. Debates in parliament were often vigorous, and legislators were usually allowed to argue and vote against the government without fear of retribution. Although the palace and the cabinet set the agenda, parliament often managed to influence policy. And this pluralism extended to Iraq's press: prior to the 1958 revolution that toppled the monarchy, 23 independent newspapers were published in Baghdad, Mosul, and Basra alone.

Not that the Iraqi kingdom always refrained from electoral fraud, harassment of opponents, or abuse of emergency powers. The government also occasionally banned newspapers that dared to indulge in particularly virulent criticism of the regime (although the bans typically lasted for only short periods). To be sure, Iraq's history -- both under the monarchy and especially after the 1958 coup -- has been filled with plenty of authoritarianism, tribalism, and ethnic and sectarian violence. The postwar reconstruction of Germany and Japan, however, not to mention the more recent transitions from communism in eastern and central Europe, all testify to the way in which democratic political institutions can change such attitudes in a country -- often quite quickly. Having said that, the success or failure of democracy in Iraq will depend on whether the country's new political institutions take into consideration its unique social and communal makeup. It is therefore important to start talking about specifics. What should the blueprint for a future democratic Iraq look like?


Iraq's ethnic and sectarian diversity -- the splits between Kurds, Arabs, and Turkmen, and between Shi`ites and Sunnis -- is usually seen as an impediment to building a stable democracy there. The fact is, however, that all this antagonism could serve a constructive purpose: having factions zealously check each others' power could actually promote democracy at the expense of rigid communal particularism. The trick is to work out a constitutional arrangement that makes sense of Iraq's social and cultural mosaic, transforming diversity into an agent for positive change.

For that reason, democratic Iraq must have a federal system of government. Already, the Kurds -- who have enjoyed freedom from Baghdad's control since the establishment of the northern no-fly zone -- have been adamant in demanding such a system. But all Iraqis would benefit from federalism, as the example of other current federal states -- the United States, Germany, Russia, and now the United Kingdom -- suggests.

In a federal Iraq, both Baghdad and the regions should be equal guardians of the constitution. Monitoring the rights and arbitrating disputes between these power bases should be the responsibility of a strong federal judiciary. As other federal states have shown, constitutional amendments to change this arrangement should be allowed only with the concurrence of both houses of the legislature, the head of state, and all federal units. Allowing the center to bypass the regions in amending the constitution quickly dilutes local rights and increases regional antipathy to central control -- as occurred in Russia before the December 1993 referendum imposed a new federal constitution.

Successful federal systems also divide power to raise and distribute revenues between the capital and the periphery. Central revenues can be used to redistribute resources from rich to poor regions, whereas local revenues support local economic and cultural initiatives. Such revenue-sharing arrangements are critical because power follows resources; when the central government denies regions the right to raise and spend money, it is tantamount to denying them authority. Revenue-sharing, on the other hand, can also decrease the temptation for one ethnic group to either capture the state or seek separation. That said, as in other federal states, certain strategic assets such as Iraq's petroleum must remain in the hands of the central government.

Local governments should in general have widespread control over their territories. This includes responsibility for all citizens in a given region, not just those of a given ethnicity. The now-collapsed Israeli efforts to give the Palestinian Authority control over some Arab activity in the West Bank and Gaza, while Jerusalem retained sovereignty over Jews in the territories, was a doomed formula: modern states, with their massive infrastructures, must be organized territorially and can function only in that manner. Limiting authorities to caring for their own kind only reinforces tribal, ethnic, and religious divisions, which can undermine democracy. For these reasons, any attempts on the part of Iraq's Arab elites to once again grant the Kurds autonomy -- without also giving them substantial control over their territory as a unit in the federal structure -- will likewise be doomed to fail.

Admittedly, federalism does not always satisfy the aspirations of groups bent on independence, as demonstrated by the conflicts in Northern Ireland, Kosovo, and Chechnya. At the same time, devolution of power has succeeded in stemming the rise of separatism in the other ethnic republics of Russia, in Scotland, and in Montenegro -- and could do the same for Iraq, if properly handled.

The question therefore becomes how to increase the chances that federalism will work in Iraq. To begin with, it would be a mistake to create only three ethnically or religiously based federal units: a Kurdish north, a Shi`ite south, and a Sunni center. Such a structure would only entrench current divisions and might even lead to ethnic cleansing. A far better idea would be to maintain Iraq's present administrative structure, under which the country is divided into 18 units. Keeping these provincial boundaries would serve the interests of Iraq's various communities, while avoiding inordinate emphasis on ethnic and sectarian concerns and increasing healthy political competition for resources -- even within various ethnic or religious communities.

Each of the 18 units should be allowed to elect a local government and send representatives to the upper chamber of a new parliament. Creating an upper house in parliament that -- like the German upper house or both houses of the U.S. Congress -- is based on regional representation would give regions a voice at the center, check the centralization of power, and, by providing a second set of local elites, minimize regional corruption. Such a system would be far better for Iraq than a centralized one, along the lines of France's old prefect system or that newly adopted by Russia, which lets Moscow appoint the governors of the seven new super-regions. Such centralized systems allow for enormous abuse, especially if the executive branch is not particularly devoted to the rule of law. The postwar occupiers of Iraq should therefore avoid even temporarily appointing Iraqi governors, since it may prove difficult to displace them once the occupation ends.


Executive branches of government are usually structured in one of two ways: unified in a single, strong presidency combining the powers of the head of state and the head of government, or divided between a head of state (a president or monarch) and a head of government (a prime minister). What would be the advantages and disadvantages of each for Iraq?

In strong unitary systems, presidents are normally directly elected, enjoy wide latitude in ruling by executive order, can call referendums to override legislation, may unilaterally declare states of emergency under conditions prescribed by the constitution, and usually have broad powers of political, administrative, and even judicial appointment. Such presidents are the primi inter pares of the branches of government and, depending on their performance and perceived legitimacy, can be extremely popular -- as, for example, was Charles de Gaulle in the first years of the Fifth Republic or Boris Yeltsin until 1993.

Too much, however, depends on the character of the individual president, and thus the disadvantages of such a system outweigh its benefits. In all but the strongest democracies, such systems are vulnerable to abuse, to coups d'etat by opposition forces, or to self-overthrow by sitting presidents who refuse to leave office once their terms are up. Indeed, one need only glance at the unitary presidencies of the Middle East and Central Asia to be reminded how prone they are to corruption, repression, and self-aggrandizement. Getting strong presidents to leave office is particularly difficult and, outside the West, rarely happens without the help of a coup, assassination, or natural death. In sum, then, strong presidencies can be judged less a mainstay of democracy than a blunt instrument for its demise.

One alternative system that might be proposed for Iraq, especially given its divisions between Arabs and Kurds and Shi`ites and Sunnis, is the Bosnian model: a shared presidency, in which each ethnic community receives a seat on a presidential triumvirate. Agreed to as part of the U.S.-brokered Dayton accord, this unwieldy arrangement was the price that had to be paid for an end to the fighting. It has, however, been beset by untold problems and has resulted in almost no state building. Each of Bosnia's three presidents is elected by, and therefore responsible only to, the electorate of one of the three ethnic communities. Unfortunately, this has reinforced the tendency of Bosnia's rival substate authorities to maintain the fiefdoms they built during the war, leaving leaders no incentive to cooperate. Bosnia's Serbian ministate in particular has remained entrenched and continues to act as a vassal of ultranationalists in Belgrade. Iraq's Shi`ites might likewise be tempted to form similar bonds with Iran, and Iraqi Kurds could look to their brethren in Turkey, Iran, and Syria -- rendering a Dayton-style shared presidency particularly dangerous for the country.

A weak but unified presidency, on the other hand, would avoid both the Bosnia scenario and the problem of creeping authoritarianism. In weak presidential systems, such as in the Czech Republic, Germany, Hungary, Israel, and Italy, the president is typically chosen by parliament and has limited, largely symbolic powers, such as recommending judges, approving constitutional amendments, and signing laws and treaties. In such systems, presidents typically cannot initiate constitutional changes, unilaterally call referendums, or prorogue parliament. The prime minister or chancellor, not the president, is the one who heads the government and does most of the work of a chief executive.

Splitting the executive between a weak president and a prime minister has a better chance of sustaining democracy in Iraq. This division would allow political dueling to take place within the democratic tent, and not in the Iraqi street. A prime minister chosen by, and dependent on maintaining, a majority in the lower house of a bicameral parliament would serve as an institutional buttress against presidential abuse and would keep the affairs of the state running. Meanwhile, a charismatic president, chosen by the upper house (itself composed of the elected representatives of the 18 federal units, as well as notables and professionals) would function as the symbolic figurehead of the Iraqi nation.

Another option that might work well for Iraq is restoring the Hashemite monarchy under strict constitutional limits. Because the Hashemites share the faith of Iraq's elite Sunni minority, restoration would reassure the Sunnis that the inevitable change in the balance of power will not lead to their marginalization. The monarchy also has the advantage of being well connected with tradition, which would make it a stabilizing force during a time of uncertainty and a barrier against extremism. A constitutional monarchy could become the symbol of Iraq's unity and civility and act as the custodian of its positive traditional values. A monarchy would also help reassure Saudi Arabia and the other Persian Gulf states that they would no longer face the kind of threat Republican Iraq has long posed.

Two obvious candidates for the throne would be Sherif Ali bin al-Hussein of Iraq or Prince Hassan bin Talal of Jordan. Both are cousins of Iraq's last king, Faisal II. Sherif Ali, a British-trained economist, is the current head of the Constitutional Monarchy Movement, an Iraqi opposition group. Prince Hassan, a graduate of Oxford University and the younger brother of Jordan's late King Hussein, has been a long-time proponent of greater democracy in the Arab world.

Of course, some Iraqis and outsiders will oppose a restoration of the kingdom on the grounds that monarchies are a thing of the past -- regressive and outdated. It is important to remember, however, that monarchies can actually help safeguard democracy. After all, when Spain restored its monarchy in 1975 after 40 years of Francisco Franco's rigid authoritarianism, the king served as a powerful guarantor of stability and progress. Moreover, in the Arab world today it is not the presidential systems but the monarchies -- Bahrain, Jordan, Kuwait, Morocco, and Qatar -- that are leading the way in democratic reform. Whatever the theoretical benefits of reinstating a limited constitutional monarchy in Iraq, however, it should be done only with popular support, as demonstrated through a referendum.


Talk of referendums leads naturally to the next question about postwar Iraq: If the country opts for a weak presidential system with a split executive, how should the president be chosen? By direct or indirect elections?

There would be advantages to both. Direct elections, in which the entire population votes for president, theoretically encourage rival candidates to position themselves at the center of the political debate, in order to maximize their chances of winning as much support as possible. At the same time, however, democratic theorists from the ancient Greeks through de Tocqueville have observed that direct elections favor populist and antidemocratic candidates. It is because of this fear that presidents in several new central European democracies are chosen by parliament.

In Iraq, where some 60 percent of the population is Shi`ite, direct elections could be expected to shift the political balance away from the minority (but traditionally dominant) Sunnis. Although such an outcome should not in and of itself be thwarted, direct elections should be avoided lest they allow the election of a Shi`ite president who unfairly favors the south or promotes an increased role for religion in state affairs. Selecting the president through indirect elections -- for example, making it a choice of the upper house of parliament, perhaps by a super-majority -- would, on the other hand, make the president beholden to, and dependent on, the success of another democratically chosen and regionally diverse body.

Prime ministers generally are leaders of the parliamentary majority in the lower house. The president and prime minister thus have different power bases and may come from different parties. During periods of so-called cohabitation, gridlock might increase, but so would the need for consensus politics. In countries where dire economic conditions demand strong and swift government, such a system can seem less than ideal. But democratic Iraq's major problem will not be economic hardship; the real threat will come from the concentration of great wealth in one industry (oil production) that is located primarily in the Shi`ite south. Getting all Iraqis to share this resource for the common good will be difficult, but institutions that diffuse power will have the best chance of success.

Turning to parliamentary elections, Iraq will face two key issues: how to draw the boundaries of its electoral districts, and how many members should be elected from each. During the monarchical period, Iraq was divided into 14 provinces, each of which was subdivided into electoral districts of 20,000 voters. These districts each elected a single member, and the complaint was often voiced that tribal leaders predominated at the expense of urban populations. Iraq is vastly more urbanized today, however. Thus it seems best to draw electoral boundaries in a way that will give greater influence to city dwellers. Doing so will also serve to strengthen secularist tendencies and decrease the possibility of rural and tribal domination of the lower house.

As for the second question, in Iraq's case there would be several advantages to having multimember districts (MMDS). For one thing, MMDS allow a district's diversity to be more clearly mirrored in parliament. In Iraq, MMDS could thus increase the representation of the professional middle class, as well as local minorities, including Sunnis in Basra, Christians in Baghdad, Turkmen in Kirkuk, Arabs in Kurdish areas, and Kurds in Arab regions. In addition, MMDS elsewhere have been shown to boost the representation of women in parliaments. Given that Iraqi women already boast high levels of education and professional attainment, increasing their input in government would contribute considerably to democratic stability. MMDS would also allow for tribal representation without allowing it to dominate.

Another way to ensure representation of women and minorities in Iraq would be to set aside guaranteed seats in parliament. For example, Iraq's pre-1958 monarchical constitution reserved a certain number of seats for Christians and Jews. More recent examples of set-aside seats, however, show that they cement rather than eradicate ethnic divisions. For example, the guarantee of representation to diaspora Croats in Franjo Tudjman's Croatia was enacted to ensure that other minorities would not outvote Croats. But the provision gave the foreign Croats rights without any responsibilities, reinforced ultranationalism within the ruling party, and limited the development of interethnic trust. Similarly, the collapse of the Good Friday power-sharing agreement in Northern Ireland, which included set-aside seats for Catholics and Protestants, has further underlined the need to avoid any quota system.

The next question is whether Iraq should enact a mixed system of voting, in which half the seats in parliament would be chosen by elections in MMDS as described above, and half would be chosen by party list. Under such a system, voters would get two ballots at each election, one for district representatives, and another for nationwide parties. Countries that currently feature mixed systems include Germany, Italy, Japan, New Zealand, Russia, and Venezuela. The advantage of such an approach for Iraq is that it would allow voters to have direct and personal contact with their local representatives while also encouraging the development of nationwide parties with national, rather than regional or sectarian, agendas. Usually, in order to run candidates in the party-list section of the election, parties need to have offices and a large number of registered members in most electoral districts. Using such standards would ensure that purely regional parties (which may be ethnic and sectarian) could still win individual seats in the lower house through mmd elections. But if they lacked offices and a nationwide membership, they would be ineligible to run in the party-list part of the campaign. Therefore, in Iraq, for example, Kurdish representation in parliament would be limited to seats won in MMDS unless the Kurds joined with others to form a truly nationwide party.

As for the upper house, following the example of Germany's Bundesrat, each of Iraq's 18 federal units would constitute one mmd, with the number of representatives per district determined by that district's population size. Legislators elected this way would constitute half of the upper body. The remaining half could be filled by elected representatives of a broad cross-section of society, such as tribal and regional notables and key representatives of professional associations (university deans and presidents, lawyers, the heads of women's associations, journalists, doctors, teachers, engineers, industrialists, and merchants). Allowing parliamentary representation by professional associations would stimulate and support the re-emergence of civil society so vital for democracy, while allowing tribal representation would reflect Iraq's history and traditions.

In accordance with typical parliamentary procedure in other countries, the upper house would not have the power to initiate legislation, but would be able to review and send back legislation deemed incompatible with the constitution and the federal system. The upper house could also be given a role in approving judicial appointments and amendments to the constitution. Providing federal units and organized interests a seat at the democratic table, even if it is in the chamber with lesser powers for everyday rule, would enhance constitutional continuity and deliberation.

In discussing electoral and party systems for Iraq, the question arises whether certain groups that have undermined democracy in the region -- specifically, the clergy, the military, and the Baath Party -- should be banned from politics altogether. As for the clerics, they could be allowed to enter the upper house as representatives of their communities (Muslim and Christian) provided they accepted the secular character of the constitution. After all, complete exclusion of the clergy has usually had disastrous consequences in the Middle East, whereas the United Kingdom, where bishops serve in the House of Lords, has shown that religious representatives can exercise a benign influence. As for Iraqi military officers, those not barred from office for complicity in the crimes of the past should be allowed to stand for election -- if, that is, they are retired from active service. The Baath Party, however, should be banned altogether in order to stigmatize it for its responsibility for the institutionalization of tyranny under Saddam Hussein. Officials who participated in torture and other human rights violations should be prosecuted and blocked from future participation in public life. And the thousands of midlevel Baath Party careerists should be allowed to return to political life only if they join new parties. Banning the party would also avoid the mistake made in some post-Soviet states, where reformed communist parties were allowed to keep their existing assets, thereby upsetting the level playing field necessary for the emergence of a competitive party system.


As almost all political theorists agree, a fully developed middle class is essential to an effective and sustainable democracy. Fortunately, even after 12 years of debilitating sanctions, a substantial and highly educated middle class has persisted in Iraq. Thus far, however, this group has not pushed for democratization or reform. This is partly because Iraq's middle class, like other sectors of the country's society, has been terrorized by Saddam's regime into submission and inaction. But there are other reasons.

Democratic theory holds that independent and self-sustaining middle classes create the basis for democratic civil life. This notion is highlighted when one glances back at the second half of the nineteenth century and contrasts the experiences of the United Kingdom and Germany. During this period, democracy was rapidly incorporated into the British body politic through a series of parliamentary acts that not only expanded the electorate, but also placed ever increasing limitations on the power of the monarchy and the aristocracy. In Germany's Second Reich, by contrast, political elites dominated by the nobility ruled through a strong authoritarian system, and chancellors and their cabinets were not answerable to the Reichstag. Britain's progress toward democracy was spurred by an entrepreneurial middle class largely independent of the state. In Germany, the economy was much more closely connected to the government, and the industrialization of the country occurred as a result of the alliance between the state and the traditional elites. There was no robust middle class to agitate for greater freedoms and representation.

Iraq's middle class today is more like nineteenth-century Germany's than the United Kingdom's. As in many other Arab countries, much of Iraq's middle class remains directly dependent on the state, primarily through employment in the vast bureaucracy, in state-owned industries, in military and security agencies, and in Baathist political bureaus. Moreover, thanks to Iraq's immense oil wealth, government revenues come mainly from oil sales and not taxation, which further adds to middle-class docility.

In order to stimulate entrepreneurship and strengthen the free market, postwar Iraq must begin the process of transferring resources from the public to the private sector. This shift should have an enormous political payoff: the development of a self-sustaining middle class that will become more proactive in promoting democratic institutions. A proper taxation system should also gradually be introduced; only then will the middle class demand accountability from the government.

Helping orient the middle class toward democracy is also important because it is the middle class that fills the ranks of any bureaucracy. Bureaucracies, known for their rigid adherence to legal rules and to hierarchies, are often not ideal agents for fostering democratic values. These impediments substantially increase, however, when bureaucracies are oversized and corrupt, as is Iraq's. It is therefore imperative to improve the country's civil servants, so they will not impede democratic growth.

Not unlike the rest of the Arab world, Iraq's bureaucracy today is simply a vehicle for ensuring full employment. This has resulted in allying ever larger segments of the middle class with the government, creating an abiding sense of dependence on, and acquiescence to, the state and its institutions. In order to lessen this dependency and improve efficiency, the current size of the bureaucracy needs to be reduced considerably.

Inefficiency and corruption would also diminish if entry into the civil service were based on merit alone; without regard for sect, ethnicity, or political affiliation. Making such a change would not be impossible, or even very difficult. True, in Saddam's political regime, unqualified but loyal employees were given posts of high responsibility. But the culture of merit is in fact embedded in the Iraqi consciousness. The most visible example of this is the national baccalaureate exam, which for decades has been taken at the end of high school. This exam has, in many ways, acted as a great equalizer; children of humble origins and resources who achieve high scores get a free education in the country's most prestigious colleges or are sent on scholarships abroad. Despite complaints of nepotism and influence peddling, the baccalaureate has also functioned as the main determinant of one's future career, and the great anxiety it engenders among all Iraqi high schoolers and their parents, rich and poor, testifies to its neutrality. To build on this foundation, a civil service college, modeled on the French ƒcole Nationale d'Administration (ena), could be established not only to teach administrative sciences but also to train future civil servants in democratic values and practices. In order to make the bureaucracy smaller and more selective, civil servants should also be paid more. Unlike in Afghanistan, however, this will not be beyond the means of Iraq, a country with the second largest oil reserves in the world.


For the sake of all parties involved, the American endeavor in Iraq must not end in a more agreeable dictatorship or a successor regime that promises nothing beyond greater cooperation with Washington. The United States' standing in the world rests not only on its might, but also on the democratic values that it espouses and propagates. The country and its allies therefore cannot shrink from setting Iraq on a democratic path. Not only will Arab and international opposition to regime change be assuaged if a democracy results; building democracy in Baghdad is also the best way to eliminate the threat of Iraqi weapons of mass destruction.

Restructuring Iraq's political system will be laden with difficulties, but it will certainly be feasible. At the same time, the blueprint for Iraq's democracy must reflect the unique features of Iraqi society. Once the system is in place, its benefits will quickly become evident to Iraq's various communities; if it brings economic prosperity (hardly unlikely given the country's wealth), the postwar structure will gradually, yet surely, acquire legitimacy. As is shown by the eastern European example, where ex-communist dictatorships have now lined up to join NATO and the European Union, putting in place democratic political institutions that function properly, meet the particular needs of a given society, and deliver the goods can rather quickly produce "habituation" -- that is, inculcate democratic habits in the population that become well entrenched and resilient. A democratic federal system would turn Iraq into the standard against which other Arab governments are judged, and make the country a natural ally of the West. Such an outcome would benefit everyone -- but especially the people of Iraq, who, after suffering for so long, deserve no less.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Adeed I. Dawisha is Professor of Political Science at Miami University, Ohio. His latest book is Arab Nationalism in the Twentieth Century: From Triumph to Despair.
  • Karen Dawisha is Walter E. Havighurst Professor of Russian Studies at Miami University, Ohio. Her books include the four volume Democratization and Authoritarianism in Post-Communist Societies.
  • More By Adeed I. Dawisha
  • More By Karen Dawisha