Do Muslims Vote Islamic?

Running Strong—or Also-Rans?

Our second conclusion is that the electoral performance of Islamic parties has been generally unimpressive. In the 89 elections that Islamic parties have contested during the past generation, these parties have typically received only a small fraction of the vote. In Pakistan, for example, the Jamaat-e-Islami (JI) and other Islamic parties have run in elections, on and off, for more than half a century. These parties reached [End Page 52] a high-water mark of 18 percent of seats in the national parliament in 1977, then ebbed below 7 percent over the next two decades. In 2002, a coalition of Islamic parties rebounded to garner 11 percent of the vote and 17 percent of seats—a major recovery, but still representing only a small chunk of the electorate even in an election that several major parties boycotted. These elections were not entirely free and fair, to be sure. Yet they did offer voters a chance to elect candidates from Islamic parties, and the voters mostly declined to do so. In 1977, the JI won 29 percent of the seats that it contested; in 2002, 27 percent of its coalition’s candidates won. In 2008, the JI chose not to compete, but a smaller Islamic party, the Jamiat-Ulema-e-Islam, won only 6 of the 108 districts in which it fielded a candidate.

It is worth noting that Pakistan is one country where Islamic parties (which do not include the large and powerful Pakistan Muslim League, despite its name) have actually performed above the norm. In Bangladesh, Islamic parties have participated in six parliamentary elections, notching their best showing in 1991 with 12 percent of the vote. Since then, they have struggled. In December 2008, the main Islamic party won less than 5 percent of the vote and only 2 of the 38 seats that it contested. A variety of smaller Islamic parties ran a total of more than a dozen candidates, yet failed to win even a single seat.

Across all 89 parliamentary elections of the past forty years in which an Islamic party participated, the median performance was 7.3 percent of the vote and 6 percent of the seats. If we combine the tallies of all the Islamic parties that participated in a given election, the median Islamic-party performance is 15.5 percent of votes and 15 percent of seats—not an insignificant portion, but hardly the lion’s share. The presence of Islamic parties did not substantially increase voter turnout, which further suggests the relative lack of voter enthusiasm for them. When turnout was higher, this did not boost the percentage of seats that Islamic parties won.

This performance has improved slightly over time, as shown in Figure 1 below, which displays the seat percentages won by Islamic parties (tracking vote percentages would reveal a similar pattern, but fewer data points are available). Algeria’s Islamic Salvation Front (FIS) was the high point in 1991, followed by Turkey’s AKP (2002 and 2006), Hamas (2006), Hezbollah in Lebanon (2005 and 2009, counting only the seats reserved for Shias in the country’s complex constitutional scheme), and Islamic parties or coalitions in Bahrain (2002 and 2006), Iraq (2005), and Jordan (1989). Below this small set of high-achieving Islamic parties is a much larger number of ideologically similar formations that do poorly. Moreover, of the 32 Islamic parties that competed more than once over the past two decades, four increased their representation by 5 percent or more, six decreased their representation by the same margin, and most have not changed much. [End Page 53]

With this context in mind, let us look at the handful of elections that Islamic parties have won. Most of these were the first truly competitive contests that these countries had experienced in a generation or more. In Jordan (1989), Algeria (1991), Bahrain (2002), Iraq (2005), and Palestine (2006), repressive secular regimes had long monopolized power, and Islamic parties promised a fresh start. Even under such exceptional circumstances, however, Islamic parties enjoy no guarantee of success. They won only a fraction of the vote in breakthrough elections in Yemen (1993), Indonesia (1999), Tajikistan (2000), and elsewhere. After the breakthrough, Islamic parties fared worse. In Jordan, a dramatic example, three-quarters of Islamist candidates were elected in the breakthrough election of 1989; this figure dropped to half in subsequent elections, and then to a quarter in the most recent election, which took place in 2007. In general, the more routine elections become, the worse Islamic parties do in them.

Do Freer Elections Help Islamists?

Of course, not all of these elections have been fully free and fair. In Egypt, for example, the Muslim Brotherhood has not been allowed to formally become a party—its candidates have to run as independents—and it is has been subject to state harassment and mass arrests during electoral campaigns. Yet the Muslim Brotherhood and many other Islamic parties [End Page 54] have persisted in their attempts to engage in electoral politics, despite such barriers. The poor performance of Islamic parties in these repressive contexts is in part a result of the rigged electoral system.

But in those Muslim-majority countries where elections were freest, Islamic parties performed worse. This finding forms our third conclusion, one that holds consistent across multiple measures of electoral fairness. The measures include:

• Judith Kelley’s “Quality of Elections” variable, which is based on the annual country-by-country human-rights reports that the U.S. State Department puts out. A country is coded as “representative” if the report makes make clear that elections there represent the will of the people. Countries that fall short are coded as “unrepresentative” or “ambiguous.”⁹
• Freedom House’s “political rights” scale, which is based on an annual survey of experts on each country. Using this scale, we categorized countries as having “greater political rights” if they scored in the top two of seven ranks with respect to the fairness of the electoral process, the extent of political pluralism and participation, and the functioning of government. Countries that fell short of the top two ranks we categorized as having “limited political rights.”¹⁰
• Polity IV’s “Electoral Competitiveness” scale, which is based on experts’ annual assessments of each country. Countries are categorized as “factional,” “electoral transition,” or “institutionalized” (scoring seven or higher on a ten-point scale) if political groups regularly compete for influence through elections. Otherwise, countries are categorized as “repressed,” “suppressed,” or “uninstitutionalized.”¹¹
• Our own “Electoral Irregularities” variable, which we developed for this project. These irregularities are coded as “systematic” if the Inter-Parliamentary Union’s electoral report mentions significant governmental barriers to Islamic political parties, such as prevention from registration as an official political party, limitation on the number of candidates permitted to run for parliament, harassment of candidates or their supporters, or manipulation of vote counts.¹²

Under all four of these methods of categorization, Islamic parties fared worse in freer elections than in less-free ones (see Table 1 below). According to Kelley’s coding of U.S. State Department reports, for example, the most successful Islamic party won a median of 9.8 percent [End Page 55] of parliamentary seats in freer elections, and 14.0 percent of seats in less-free elections. The median for all Islamic parties in a given election was 11.5 percent under freer conditions and 15.9 percent under less-free conditions. The Kurzman and Naqvi categories, which refer specifically to the experience of Islamic parties as opposed to parties as a whole, yield an even larger differential: The median percentage of seats won by Islamic parties in freer elections was almost 10 percentage points lower than the percentage won in less-free elections.

This pattern is visible across all Muslim-majority countries, whether they are Arab or non-Arab. Overall, Islamic parties have done better in Arab countries than elsewhere, winning an average of 15 percent more seats. At the same time, Arab countries have had fewer democratic elections than Muslim countries outside the Arab world.¹³ Within the Arab world, the median seat total for Islamic parties was several percentage points lower in relatively free as compared to less-free elections.

These findings run directly counter to concerns that free elections will lead inevitably to victories by Islamic parties. A caveat is in order, however: The freeness of elections in a given country may be directly related to the government’s estimate of how popular Islamic parties are there. If governments restrict electoral freedoms more in countries where Islamic [End Page 56] parties seem to be popular, then Muslim societies with free elections would tend to be places where Islamic parties are less popular.

A further caveat has to do with the practice of running partial electoral slates. In a number of countries with semi-free elections, Islamic parties run candidates only in a limited number of districts out of fear that running in more will make the government nervous and lead to stepped-up repression. Just before the 2003 voting in Jordan, for example, a leader in the Islamic movement noted that his party “had assured the palace that they did not seek to gain a majority in the upcoming elections.”¹⁴ In Lebanon in 2000, Hezbollah might have won several more than 11 of 27 Shia seats had the Syrian government not required it to limit the number of candidates that it fielded.¹⁵ At times it may be hard to know whether partial slates reflect government pressure or an Islamic party’s own choice to focus on districts where it has the best chance. In Malaysia, for example, the Pan-Malaysian Islamic Party contested 46 of 192 seats in 1995 (winning 7); 63 of 193 seats in 1999 (winning 27); 84 of 219 in 2004 (winning 7); and 68 of 222 in 2008 (winning 23).¹⁶

Liberalizing or Radicalizing?

Among Islamic parties that have joined national elections, there has been a trend toward the adoption of more liberal electoral platforms, as illustrated in Figures 2 and 3 below. Prior to the mid-1990s, a majority of the platforms in our collection favored the implementation of shari’a and a ban on interest and made some mention of jihad and opposition to Israel. Since that time, half or fewer of the platforms since have adopted such positions. By contrast, recent platforms are more likely to mention democracy, the rights of women, and the rights of minorities. Hence our fourth and final conclusion: Islamic parties have (relative to their starting point) liberalized their stances significantly over the past several decades.

Looking at the twelve Islamic parties for which we can construct a “platform timeline” of sorts—that is, two or more platforms from different dates—we can see slight shifts in a liberalizing direction. Four of these parties dropped their support for shari’a (for the record, they were the Renaissance Movement in Algeria, the Islamic Action Front in Jordan, the Jamaat-i-Islami in Bangladesh, and the Justice and Development Party in Morocco). No party added a call for shari’a to its agenda. Eight parties maintained their position across time, and split evenly: Four continued promising to implement shari’a, and four continued to offer no such pledge. On the issue of democracy, three parties switched—and all moved toward offering secular justifications for democracy. Three of five parties that referred to jihad in their platforms removed this reference in later platforms (they were Egypt’s Muslim Brotherhood, the Jamaat-i-Islami in Bangladesh, and the Yemeni Congregation for Reform).

By any standard, these are still highly conservative parties, and there [End Page 57] are occasional examples of movement away from liberalization as well. The most widely noted case concerns the Egyptian Muslim Brotherhood. In August 2007, it circulated a new platform draft that proposed, for the first time, naming a council of religious scholars whose mission would be to advise the government regarding the compatibility of its programs and policies with shari’a. In some cases, moreover, this council would have the authority to overrule state decisions on grounds drawn from Islamic law.

This marked a departure from the Brotherhood’s earlier platforms, which had always proposed allowing current government bodies to make decisions concerning compatibility with shari’a. In its 2005 platform, the Brotherhood offered the general principle that “there is no one in Islam who has religious authority—whoever he is—except the authority of good preaching, calling for good and warning of evil.” By advocating the creation of an authoritative religious body, the 2007 draft rescinded this principle. At the same time, the 2007 draft—which the Brotherhood never formally adopted, perhaps because of the outcry that it sparked—waxed more liberal on other issues. It called for stronger support of women’s rights, for instance (though here too, its recommendation that women be kept ineligible for the presidency roused controversy).¹⁷

Notwithstanding the exceptions, the Islamic parties’ overall trend toward publicly embracing global norms of democracy and human rights is significant. It is visible not just at the party level, but also in the ideals and behavior of individual activists. As political scientist Carrie Rosefsky Wickham has noted, many leading Islamic politicians seem to be changed men—two decades ago, it was hard to find one who would shake a woman’s hand or engage in interfaith dialogue, but today many do. The experience of political participation, both in government and in [End Page 58] civil society, has changed their outlooks in ways that they did not imagine when they started down the path of electoral politics.¹⁸

One mechanism for this transformation is the strategic goal of appealing to swing voters. As they compete with secular parties and rival Islamic formations, Islamic parties can either resign themselves to fringe status or move toward the center of the ideological spectrum. This center, in most Muslim societies, is culturally conservative and politically liberal. According to the World Values Survey, which has polled more than eighty countries over the past decade, Muslims are just as likely to support democratic ideals as non-Muslims, but are considerably less comfortable with abortion, divorce, and the employment of mothers of young children. The survey’s Muslim respondents overwhelmingly favor multiparty elections, and they also believe that political leaders should be inspired by religious values.¹⁹ Islamic parties are not alone in targeting this median position. Many secular parties in Muslim societies—and in many other societies besides—appeal to constituencies of culturally conservative democrats.

Yet popular religiosity may not be the most significant factor shaping the performance of Islamic parties. In Pakistan, for example, 62 percent of the adult population believed in 2001 that “good government … should implement only the laws of the shari’a.” In Bangladesh, 44 percent of respondents said the same. Yet few voters supported the Islamic parties that claimed to share this goal. In Egypt, 80 percent of respondents favored implementation of shari’a in a 2000 survey; that same year, a only quarter of the Muslim Brotherhood’s candidates won election to parliament. Five years later, with surveyed support for shari’a declining slightly, 61 percent of the Brotherhood’s candidates won.²⁰ In [End Page 59] this and other instances, Islamic-party success seems to have been based on factors separate from popular support for the implementation of Islamic law. For example, an exit poll by the Palestinian Center for Policy and Survey Research showed that Hamas’s electoral victory in 2006, with 44 percent of the vote, was due far more to voters’ concerns about corruption and the lack of security than to religiosity.²¹

Throughout the past generation, Islamic parties have faced resistance from Islamic revolutionaries who object to electoral politics. Most famously, al-Qaeda’s leaders have consistently denounced voting as a human usurpation of divine authority. Two decades ago, even before allying himself with Osama bin Laden, Ayman al-Zawahiri wrote a tract condemning the Muslim Brotherhood’s abandonment of revolutionary methods in favor of electoral politics, equating this decision with apostasy: “Whoever labels himself as a Muslim democrat, or a Muslim who calls for democracy, is like saying he is a Jewish Muslim or a Christian Muslim.”²²

When Hamas decided to participate in the Palestinian elections in 2006, Zawahiri issued a series of statements that called the Palestinian Islamists sellouts.²³ During the 2008 elections in Pakistan, the local Taliban accused the country’s Islamic parties of seeking to enrich themselves with the perks of office.²⁴ In several countries, Islamic-party leaders have faced violent attacks by militants. During the bloody civil war that wracked Algeria in the 1990s, dozens of leaders of the Movement for an Islamic Society (now the Movement for a Peaceful Society), which favored a peace settlement, were assassinated—some by Islamic revolutionaries and others by state-security forces. In Iraq, Sunni Islamic revolutionaries have recently renewed their campaign “to start killing all those participating in the political process,” according to a warning received by a Sunni politician who was subsequently assassinated in Mosul.²⁵

The proliferation of Islamic parties in the electoral process—despite unfair conditions, threats from revolutionaries, and a generally poor record of vote-getting so far—is a sign of how far such parties have gone in embracing liberalization. A quarter-century ago, many of these movements sought to replicate the Iranian Revolution and create an Islamic society by overthrowing the state. By the end of the twentieth century, however, most Islamic activists found themselves disillusioned with the Islamic Republic of Iran and other Islamic states such as Sudan or Taliban-run Afghanistan. Even in Iraq, where a majority of Shias voted for Islamic parties in the 2005 parliamentary elections, less than a quarter of [End Page 60] Shia survey respondents said that they favored a theocratic system like Iran’s. Only 6 percent of Iraqi Sunnis favored such a system.²⁶
Instead, Islamic parties may be reprising an earlier historical moment, the watershed period of the early twentieth century when demands for democracy and human rights first gained mass support in Muslim-majority societies. In 1905, Muslims of the Volga Basin, the Caucasus, and Central Asia joined the rest of the Russian Empire’s people in agitating for a representative assembly. In 1906, Iranians staged large sit-ins to force the shah to accept a constitution and an elected parliament. At its first meeting, in 1907, the All-India Muslim League unanimously adopted a resolution favoring elections under the British colonial regime. In 1908, Ottomans celebrated en masse when the sultan was forced to reinstate the constitution and schedule parliamentary elections. Many Islamic leaders participated actively in these movements. The Shaykh al-Islam, the Ottoman Empire’s highest-ranking Islamic official, sided with the democracy movement and against the sultan. The two most influential Shia scholars of the era sent telegrams calling opposition to constitutionalism un-Islamic.²⁷

Then, as now, reactionary religious forces resisted the entry of Islamic parties into electoral politics. But that was not what ultimately undermined the democratic trend in early twentieth-century Muslim societies. Rather, it was secular authoritarians—the Committee of Union and Progress and later Mustafa Kemal in the Ottoman Empire, the Bakhtiyaris and later Reza Khan in Iran, and European colonial authorities elsewhere—who dismissed parliaments, shuttered newspapers, and suppressed prodemocratic Islamic movements.

It would be premature to suggest that this history is about to repeat itself. But one aspect of the story already looks like a rerun. Just as they did a century ago, authoritarian states and Islamic revolutionaries alike are striving to keep Islamic political parties from competing freely for votes. The authoritarians have succeeded in one sense: Islamic parties have won very few elections. In a broader sense, however, the efforts to suppress these parties are failing. The evidence suggests that suppression of Islamic electoral options only makes them more popular with voters. According to the World Values Survey, support for shari’a is highest in the countries with the lowest levels of political freedom, such as Egypt and Saudi Arabia.²⁸ By contrast, when Muslims are given the opportunity to vote freely for Islamic parties, they have tended not to do so. [End Page 61]