Intersectional Representation Between Gender, Religion, and Nationality

Immigrants who came to Europe in recent decades (work immigrants and/or refugees) grapple with intersectional identities, such as religion, nationality and gender, yet current political research addresses these issues only in part. To address these omissions, I conducted a content analysis of all parliamentary questions Muslim representatives raised in their parliamentary activities in three Western countries. I also investigated whether the representatives' invisibility pertains only to their descriptive representation or whether it affects their substantive representation by analyzing five research hypotheses for differences in the content of the parliamentary questions. I found that male and female Muslim representatives ask parliamentary questions about Muslim men and women. In addition, I developed an Intersectional Representation Index to measure and demonstrate the complexities Muslim representatives face in Western countries. The index shows that such representatives have several identities, some of which have become invisible, as previous studies indicated.


Introduction
Scholars have argued that little research attention has been given to the political representation of Muslim minority women in the West (Purdie-Vaughns and Eibach 2008;Ruedin 2013;Hughes 2016). Furthermore, most of the existing research concentrates on descriptive representations of Muslim women, thereby ignoring two components of the concept of representation: 1) the issue of gender regarding Muslim representatives and 2) the substantive representation of both Muslim men and women.
Immigrants who came to Europe in recent decades (work immigrants and/or refugees) grapple with multiple identities, such as religion, nationality, tradition, and gender, yet current political research addresses these issues only in part. Analyzing just one or two identities, as previous studies did, might provide just a partial picture of the challenges representatives who have intersectional identities face. My goal is to combine these different identities into one overall concept of intersectional representation by discussing how each identity is reflected in the content of parliamentary questions that male and female Muslim representatives ask. I also examine when some identities disappear and others move to the forefront. I maintain that intersectional representation leads to greater invisibility among representatives who belong to several disadvantaged groups and have several identities, compared with those who belong to only one disadvantaged group. Furthermore, I compare the Muslim representatives in the UK, Germany, and Israel. In the first countries, all the representatives are immigrants or the children of immigrants. However, in Israel, all the Muslim representatives are native-born Israelis. My results indicate that while Muslim representatives in Israel are more coherent regarding their Muslim identity, Muslim representatives in Germany and the UK address religious issues in their parliamentary questions but have other identities to represent as well. Furthermore, Muslim women's issues are generally invisible in the UK and Germany, whereas in Israel they are addressed more often.
There are just a few other related papers that use data on the same legislative mechanism--parliamentary questions-- (Saalfeld 2011;Saalfeld and Bischof 2012;Kolpinskaya 2017) to study the same area --religion in politics (Saalfeld 2011;Saalfeld and Bischof 2012;Kolpinskaya 2017). However, all of them analyzed only one country --the UK. Furthermore, they did not address the main issue of intersectional representation. Thus, my analysis contributes to the literature by adding two more countries (Germany and Israel), and gender as an additional identity along with religion to the analysis and the concept of intersectional representation. The previous studies explored whether the growing descriptive representation of ethnic minority representatives in the UK parliament has any implications for the substantive representation of minority-related issues in the UK Parliament (Saalfeld 2011). They mapped and explained variations in the extent to which Black, Asian and ethnic minority representatives use written parliamentary questions to articulate issues relevant to minority constituents (Saalfeld and Bischof 2012). They also investigated the factors that drive UK representatives to raise issues of concern to Jewish and Muslim minority groups using written parliamentary questions (Kolpinskaya 2017). They found that minority and non-minority representatives alike ask more questions related to minority concerns, if they represent constituencies with a large share of non-White residents. However, representatives with a visible minority status do tend to ask significantly more questions about ethnic diversity and equality (Saalfeld 2011). Finally, a background as a member of a religious minority has a limited impact on the representatives' engagement with minority issues in written parliamentary questions (Kolpinskaya 2017). I argue that analyzing just one or two identities, as the mentioned studies did, might provide just a partial picture of the challenges representatives who have intersectional identities face. Hence, I maintain that intersectional representation leads to greater invisibility among representatives who belong to several disadvantaged groups and have several identities, compared with those who belong to only one disadvantaged group.
The article proceeds as follows. First, I present the previous research on intersectional representation and parliamentary questions as a representative tool. Second, I describe the characteristics of the Muslim populations in the three countries. Third, I explain the methodology of the paper. Fourth, I report the results of the content analysis and discuss the findings. Finally, I explore the implications of the findings in the conclusion section.

Intersectional Representation
The literature on intersectionality distinguishes between multiple and intersectional identifications (Hancock 2007;Mügge and Erzeel 2016;Geese and Schwemmer 2019;Mügge et al. 2019;Reingold et al. 2019). Hancock (2007) asked six questions to distinguish between the unitary, multiple and intersectional approaches. She argued that intersectionality emerges out of the earlier unitary and multiple approaches. Based on the six questions, she indicated that the intersectional approach involves the dynamic interaction between individual and institutional factors, and that members often differ in politically significant ways. In contrast, the other approaches are static at the individual or institutional level, and members are uniform in their political methods.
Previous studies found that individuals who belong to two or more disadvantaged groups are more invisible, in terms of political representation, than those who belong to only one disadvantaged group (i.e., are "doubly disadvantaged") (Strolovitch 2006;Purdie-Vaughns and Eibach 2008;Celis and Erzeel 2017;Hancock 2016). Such studies, however, analyzed only part of the process of political representation-becoming a candidate and getting elected (Hughes 2016), examining the descriptive representation that Pitkin (1967) defined as the true reflection of the essential characteristics of the voters in terms of their gender, race or ethnicity. For example, Hughes (2016) found that Muslim minority women are increasingly elected in countries with proportional representation electoral systems, whereas Muslim minority men are customarily elected across a range of electoral systems. Purdie-Vaughns and Eibach (2008) argued that people with multiple subordinate group identities are defined as non-prototypical members of their respective identity groups and experience "intersectional invisibility." Do these representatives lose their multiple identities and concentrate on the one they think is more relevant or is the definition of "intersectional invisibility" made by other representatives and/or the public, while representatives with multiple subordinate group identities continue to act in all of their identities? In other words, do these representatives adopt an intersectional approach and hence have dynamic interactions between individual and institutional factors, and are different in politically significant ways?
Recent studies on the descriptive representation of minorities found that minority women are still substantially underrepresented in national legislatures (Hughes 2013a;Zingher and Farrer 2016). In Belgium, political parties prefer ethnic minority women candidates because their "intersectional identity mix" is maximally complementary to groups embodied by the incumbents (Celis and Erzeel 2017). Hughes (2013aHughes ( , 2013b investigated how women's political representation varies by race, ethnicity, and religion at the global level. She argued that by focusing on them as a group, we might overlook the different outcomes of women who are marginalized by their minority status, missing even greater levels of underrepresentation of some groups of women around the world. Unsurprisingly, Hughes found that minority women are substantially underrepresented in national legislatures overall, but their level of exclusion varies geographically. Indeed, there has been increased interest in the study of intersectionality (Htun and Ossa 2013;Mügge and Erzeel 2016;Geese and Schwemmer 2019;Mügge et al. 2019;Reingold et al. 2019), and increased awareness about the need for race and gender researchers to work together.
Although descriptive representative is important, my main goal is to analyze substantive representation, which Pitkin (1967) defined as the actual activities of representatives in their daily work in the House or in their constituency. Mugge and Erzeel (2016) identified it as the third step in the mechanisms of the representation process, the in-office representation of representatives' ideas. Previous studies determined that immigrant-origin representatives ask more migration-related parliamentary questions and are more likely to be concerned about the substantive interests of minorities in the parliament (Saalfeld and  My study differs previous ones in several ways. First, since the previous studies were conducted, the number of immigrant-origin representatives in Germany has increased, and representatives already in place have become well-established. Second, I added two countries-Israel and the UK-to the analysis to explore similarities and differences in the substantive representation of immigrant-origin representatives in Germany. Furthermore, Israel is unique because its Muslim representatives are native-born. 1 In contrast, in previous studies in the UK and Germany these Muslim representatives are immigrants. Third, and most importantly, I analyzed only Muslim representatives (men and women), while previous studies considered immigrants from all countries and all religions (Wüst 2014). Murray (2016), who analyzed France as a case study, argued that the content of legislation (such as the 2004 legislation that prohibited all prominent religious symbols in state schools) may fulfill the substantive interests of some minority (religious) women, on one hand, while compromising the civil rights of others (secular and liberal women).
I argue that representatives with intersectional identities must address the various aspects of these identities in their representation, which I define as intersectional representation. For example, Muslim women representatives in Western democracies are expected to deal with issues related to the Muslim minority in their countries, women, and the issues of concern to their party and constituency (Saalfeld and Bischof 2012).
Scholars agree that representation is a complex concept, both theoretically and empirically. They also agree that representation can be defined by the continuing responsiveness of the representatives to voters' preferences (Hughes 2013b), and that responsiveness is both descriptive and substantive (Pitkin 1967). Attempts to define representation roles using dichotomous and ideal types do not represent reality. Therefore, how do we measure intersectional representation? I argue that when representatives have intersectional identities, they have intersectional representation roles. To understand the concept, I will start with some basic assumptions. First, political actors are rational actors, as previous studies indicate (Hug 2014;Fox 2018). Second, as such, they are familiar with different kinds of representation typologies. Third, they realistically decide which representation typology to adopt based on the relevant identity they wish to represent at that time.
The concept of intersectional representation is like previous studies that focus on representatives' self-perceptions rather than on social expectations. Furthermore, the concept of intersectional representation embraces Pitkin's (1967) claim that political representation is not an atomized relationship between citizen and representative but rather a dynamic, daily, ongoing activity to meet the challenges of addressing intersectional identities. It is not, however, an institutionalized arrangement between many citizens and groups as Pitkin claimed, but rather an ongoing struggle between representatives, parties, identities, and citizens.
Several recent studies explored the behavior of immigrant-origin minority representatives in European countries (Saalfeld 2011;Saalfeld et al. 2011;Wüst 2014). In contrast, I focused on Muslim immigrants due to the new wave of such immigrants in Europe. I wondered whether Muslim representatives even represent newly arrived Muslims. The first Muslim representatives were elected to the British and German parliaments in 1997 and 1998, respectively, and most do not share the same characteristics as the new Muslim immigrants in these countries. The earlier arrivals are more educated, pursue liberal professions, have fewer children, are secular, and hold liberal opinions, whereas current Muslim immigrants are religious, less educated, have more children, and are more conservative in their culture and views. As Murray (2016) argued, ethnic minority women who have made it into politics are highly assimilated and secular. As such, they are symbols of the Universalist French Republic but do not represent the full diversity of women of immigrant origin. In Israel, the Muslim minority identity is partially Arab (nationality), partially Palestinian (identity), and partially Israeli (citizenship). Therefore, they have three identities-Israeli, Arab and Palestinian-in addition to their religious identity as Muslims.
In intersectional identities, nationality plays significant role. People can live in one country and feel that they belong to another country. This is one of the challenges immigrants face. Scholars refer to nationality and citizenship as the formal status of state membership and offer several definitions based on different contexts. Evidence from previous studies indicates that citizenship faces more challenges and complexities in divided societies that contain conflicting national or cultural identities. Minority groups might feel that their cultural identity and distinctiveness are threatened when there is a superordinate identity (Ben-Porat and Turner 2011; Bloemraad and Sheares 2017;FitzGerald et al. 2018).
Israel, Germany and the UK face different challenges regard their notions of nationality. Israel has long been viewed as the prototype of an immigrant society, having the highest proportion of foreign-born citizens of any country (Semyonov et al. 2015). Thus, while Jews and Arabs formally enjoy equal citizenship rights by law, they vary substantially in practice (Ben-Porat and Turner 2011).
The current public debate on the integration of Muslim immigrants in Europe revolves around the desire for a common European identity, which is defined in terms of shared political and moral values, rather than a national culture, history or language. Scholars have argued that it can create a dangerous moral superiority because it implies that immigrants have to support these universal values even when they oppose them (Mavelli 2013;FitzGerald et al. 2018).
Given these intersectional identities, I ask several questions. Do Muslim representatives substantively represent Muslims? How do they cope with all the other identities they have, or do they fulfill a symbolic role only? I answer these questions by analyzing the content of parliamentary questions asked by Muslims representatives.

Parliamentary Questions as a Representative Tool
Scholars consider parliamentary questions an easy parliamentary procedure tool to use (Martin 2011;Jensen et al. 2013;Wüst 2014). In addition, the time required to formulate and ask questions is very short. The procedures are immediate, and the main benefit is that representatives can force ministers to check or reveal information, policies, and actions they do not always wish to share. While some questions genuinely aim to seek information or action, others are designed to highlight the alleged shortcomings of the minister's department or the merits of an alternative policy. Representatives also use written questions to extract more detailed information from the government than would be available in response to an oral question.
Scholars have found that minority representatives use parliamentary questions more frequently because they are a less constrained parliamentary tool (Saalfeld 2011;Kolpinskaya 2017). Indeed, representatives with immigrant backgrounds use the tool to emphasize the unique challenges immigrant citizens face and to ask more migration-related parliamentary questions (Saalfeld, 2011;Saalfeld et al. 2011;Wüst 2014). Saalfeld et al. (2011) analyzed the content of written parliamentary questions with the objective of answering questions about the descriptive and substantive representation of immigrant-origin parliament representatives in Germany, France and the UK. Such questions included, what types of representatives with migrant or visibly minority backgrounds get into parliament? What do such representatives do as representatives once they are elected? Do they focus on specific roles, and why do they assume these roles? Saalfeld and Bischof (2012) similarly analyzed parliamentary questions posed during the 2005-2010 UK parliament and the first 20 months of the following parliament (May 2010-December 2011). They found that ethnic minority representatives are more likely to address the substantive interests of minorities, but that it is the represented party and the constituencies' ethnic diversity that drive this form of representation.
Parliamentary question procedures differ slightly from parliament to parliament. The UK, Israeli and German parliaments all have three types of parliamentary questions. The Bundestag has major, individual, and minor parliamentary questions (interpellations), 2 the Israeli Knesset has written, oral, and direct parliamentary questions, and the UK parliament has oral, written, and urgent parliamentary questions.
I provide a four-fold classification of parliamentary questions that is based on two variables-the main arena for the answer (on the floor or written answers) and who is entitled to ask the question (a faction or individual representatives). This classification helps us understand the strength of parliamentary questions as a tool for overseeing the government. Individual representative As the table shows, the first main difference between Germany on one hand and the UK and Israel on the other is the importance of the faction versus the individual representative. The main two powerful parliamentary questions in Germany (major and minor) are in the hands of the faction 5 , while in the UK and Israel they are in the hands of individual representatives only. Another difference is in the arena of the answer: both major and individual parliamentary questions in Germany can be answered on the floor or as written answers, as per the asker's request. In the UK and Israel, each type of parliamentary question has its own arena in which it is answered. Representatives cannot choose to receive the answer in more than one defined arena. 6 I analyzed major and minor parliamentary questions in the German parliament and written parliamentary questions in the UK and Israeli parliaments for the following reasons. First, oral (UK and Israel), urgent (UK), and individual (Germany) parliamentary questions are used less compared with the types of parliamentary questions I chose. Since there are relatively few Muslim representatives in these parliaments, I chose to analyze a larger number of parliamentary questions for each representative. Second, party (hence, the faction) matters, especially in the European context, so I explored whether representatives are bound by directives dictated by their party or their constituency. In the German parliament, the faction is much more significant in the daily life of individual representatives than in Israel and the UK. It is interesting to note that some Muslim representatives in Israel are members of Jewish political parties while others are members of Arab parties (a yet additional identity). In Germany, most Muslim representatives are members of the left wing, liberal Green party. In the 2015 UK elections, thirteen Muslims representatives were elected, eight of them women (gender identity); nine were from the Labour Party, three from the Conservative Party, and one from the Scottish National Party. In all of these countries, the party affiliation adds to the various identities a single representative has. Thus, comparing parliamentary questions by German representatives with those by British and Israeli representatives can reveal differences in the factions' power in these parliaments.
Hence, I posited several hypotheses: H1: Parliamentary questions by Muslim representatives will contain more Muslim issues than parliamentary questions on all issues other than Muslim issues.
H2: Female Muslim representatives will ask more parliamentary questions on women's issues compared with male Muslim representatives.
H3: Female Muslim representatives will ask more parliamentary questions on Muslim women's issues compared with male Muslim representatives.
H4: The three countries will differ in the average number of Muslim issues raised.
In addition to the explanatory variables of religion, gender and nationality, previous studies considered additional variables. I added another one-membership in the ruling coalition or the opposition-because it explains l representatives' behavior in general, and their use of parliamentary questions in particular. When representatives are part of the majority party, they are better able to advance their agenda or achieve their goals than when they are part of the opposition (Jensen et al, 2013). Opposition representatives, however, often use parliamentary tools that have few restrictions to be heard (Martin 2011;Kolpinskaya 2017).
Hence, my final hypothesis was: H5: Opposition Muslim representatives will ask more parliamentary questions addressing Muslim issues than coalition Muslim representatives.

Muslims in Germany, the UK, and Israel
To test the research hypotheses, we first must understand the characteristics of the Muslim populations in these countries. 7 And the number of Muslim representatives and Muslim women representatives in these countries as part of descriptive 5 As described in the Rule of Procedure of the German Bundestag "Items of business submitted by Members of the Bundestag (Rule 75) shall be signed by a parliamentary group or five per cent of the Members of the Bundestag" (rule 76, article 1). For more information see https://www.bundestag.de/service/glossar#url=L3NlcnZpY2UvZ2xvc3Nhci9nbG9zc2FyL0EvYW5mcmFnZW4vMjQ1 MzI0&mod=mod445382 https://www.bundestag.de/service/glossar#url=L3NlcnZpY2UvZ2xvc3Nhci9nbG9zc2FyL0sva2xlaW5lX2FuZnJhZ2Uv MjQ1NDc2&mod=mod445382 https://www.bundestag.de/service/glossar#url=L3NlcnZpY2UvZ2xvc3Nhci9nbG9zc2FyL0cvZ3Jvc3NlX2FuZnJhZ2U vMjQ1NDUy&mod=mod445382 6 http://archive.ipu.org/parline-e/Modlist.asp representation. 8 According to Israel's Central Bureau of Statistics, the Arab population is currently (2019) estimated at 1,907,000, representing 20.7% of the country's population. Most of them identify themselves as Arab or Palestinian by nationality and Israeli by citizenship. Most of them are Muslims (83.8%) and the rest are Christians, Druze and Circassian. Since the establishment of the State of Israel in 1948, all Israeli Arabs are Israeli citizens with the right to vote and to be elected.
Israel has had an increasing number of Muslim legislators since its establishment, but only in the 9th Knesset in 1977 were Muslim representatives part of an Arab party rather than part of an Arab satellite list 9 or a Jewish party (see Figure 1). Nevertheless, the descriptive representation of the Muslim minority in Israel is less than their share of the total population. Although the descriptive representation of Muslim women has increased, it has done so in an even more underrepresented manner than for Muslim representatives in total or male Muslim representatives. The 2011 UK census revealed there were roughly 2.7 million Muslims living in Britain, 4.83% of the overall population. The largest Muslim population was of Pakistani origin, followed by Bangladeshi origin. Other Muslims were from South Asia and other countries like Yemen, Syria, Iraq, Lebanon, and Jordan. There was also a growing number of indigenous British Muslim communities consisting of white converts to Islam. Furthermore, each group can be split into first, second and now third generation Muslims, all having different experiences of what it means to be and live as a Muslim in Britain today.
With approximately 3.0 to 3.5 million Muslims, Germany currently has the largest Muslim population in Western Europe after France (4.3% of the population and growing continuously 10 ). Eighty percent of these Muslims are not German citizens (only about 608,000 are, and 100,000 of those are German converts to Islam). Seventy percent of the Muslim population is of Turkish origin, constituting the largest and most dominant national minority in Germany today.
Since most Muslims in Germany are not German citizens, they are denied the right to vote or actively participate in the political sphere. Nevertheless, although fewer Muslims are taking up German citizenship, children are being born as German citizens, even when their parents are not German citizens. 8 The information on the religion of representatives came from various sources, including parliament websites, Wikipedia, representatives' personal websites and http://www.euro-islam.info. All data were cross-checked. Representatives were coded as "Muslim minority" if there was public information stating their background as such.  1949-1951 1951-1955 1955-1959 1959-1961 1961-1965 1965-1969 1969-1974 1974-1977 1977-1981 1981-1984 1984-1988 1988-1992 1992-1996 1996-1999 1999-2003 2003-2006 2006-2009 2009-2013 2013-2015 2015 Total Figures 2 and 3 present the number of Muslim representatives in the UK and Germany. In both countries, the first Muslim representative was elected just before the end of the last century (1997 and 1998, respectively). Since then, the number of Muslim representatives and of Muslim women representatives in particular has increased in both countries. Unlike in Israel, however, there are no Muslim parties in their parliaments, so they all belong to existing parties. Three female Muslim representatives entered the UK parliament for the first time in 2010. Six of the elected representatives were members of the Labour Party and two were the first Muslim Conservative representatives ever. In the German parliament, most of the Muslim representatives are from the Green Party, which focuses on ecological, economic, and social sustainability issues. The rest are members of the Left Party and the Social Democrat Party.  Vol. 13, No. 4;2021 than in Israel. 11 The literature provides several explanations for this disparity, the first of which involves the local political culture. For example, Nordic countries all rank among the highest in the proportion of women representatives. Two indicators from the OECD dataset on women show that Israel is less favorable toward women than Germany and the UK. The first is the OECD gender wage gap, 12 defined as the difference between the median earnings of men and women relative to the median earnings of men. According to this indicator, the wage gap in Israel is 19.3%, compared with 16.8% in the UK and 15.5% in Germany. The second is the OECD women in politics indicator 13 , defined as the proportion of women in the lower or single house of parliament. Again, Israel is less favorable toward women compared with the UK and Germany (27.5%, 30% and 37%, respectively). Furthermore, most Muslim citizens in Israel live in predominantly non-Jewish villages and cities. Hence, their traditional society is a significant factor that affects their incentives and abilities to participate in political life (Ghanem and Mustafa 2011;Jamal 2011).
The second explanation is that electoral institutions are an important factor affecting women's representation levels. Hughes (2016) found that Muslim ethnic minority women are increasingly elected in countries with proportional representation electoral systems in which parties include viable female candidates on the party list in order to appeal to a broader electorate. Large district sizes in multimember systems mean lower thresholds to cross when seeking nomination. Furthermore, centralized control over nomination procedures means that party elites can increase the number of viable women candidates in response to pressure for greater representation. In contrast, most single-member district systems tend to have decentralized nomination structures. Indeed, Israel has a proportional representation electoral system but without constituencies. That fact, coupled with the dominance of tradition in Arab society, makes it difficult for Muslim women to enter politics. All four Arab women elected to the Israeli parliament were born in cities (which are less traditional), two of which are mixed, more liberal cities. Also, two of the women were elected through Jewish parties that have quotas for women and Arabs.

Methodology
The research dataset covered the total activity of each Muslim representative (excluding Muslim representative who were ministers or deputy ministers 14 ) during one parliamentary term in each country. Hence, for the UK, I analyzed written parliamentary questions for 2015-2017 (56 th parliament), for the Germany parliament, I examined the 2013-2017 term (18 th Bundestag), and for the Israeli Muslim representatives, I analyzed parliamentary questions for 2013-2015 (19 th Knesset term). In total, I analyzed twenty-five Muslim representatives: five from Germany, eleven from the UK, and nine from Israel. Fifteen (60%) were male and ten (40%) were female; twenty-one (84%) were senior representatives, while only four (16%) were junior representatives; twenty-two (88%) were from the opposition and only three (12%) were from the coalition (all of them in the UK).
The content analysis I conducted of the parliamentary questions was based on grounded theory, which served as the underlying framework for the open coding and axial coding of the interviews. The process involves interpreting the codes, categories, and properties developed in open coding to refine the constructs and make them more abstract and theoretical. The first step was based on Wüst's (2014) definition of migration-related issues as pertaining to "immigration (not emigration), migrants, foreigners, integration, discrimination (also: right-wing extremism; but not: women per se) and minorities. Most of these issues can be considered domestic, whereas some also concern EU politics or discrimination in foreign countries. Questions referring only to foreign affairs such as discrimination against minorities in other countries, for example, in Afghanistan, India or Turkey, are not coded as migration-related" (p. 9). Thus, I defined Muslim issues as those pertaining to integration and discrimination, as Wüst did. I also added two issues of concern to Muslims: employment and civil rights as citizens. Issues such as civil rights in Muslim-majority countries were not considered Muslim issues within the country but were coded as Muslim issues outside the country.
The categorization method used two levels. The first classified parliamentary questions by 1) gender, 2) religion, and 3) nationality, and the other consists of 20 topics. In order to overcome the potential overlap between categories and to code properly, I employed two research assistants, one in each country (Germany and the UK 15 ). Each of them applied the coding I defined previously for 100 parliamentary questions, which I too coded separately. The content analysis of the first 100 parliamentary questions yielded over twenty new themes in addition to Muslim issues and Muslim women's issues. Since there were some discrepancies between the two datasets, I analyzed each discrepancy to understand the research assistants' decisions. After all discrepancies were resolved, each research assistant coded the remaining parliamentary questions, including the additional themes found in the preliminary coding process such as nationality (Turkish for Muslim representatives from Germany and Palestinian for Muslim representatives from Israel), or general issues that were referred to frequently by several representatives in their parliamentary questions, including foreign policy, refugees and immigrants, policy toward third world countries and civil rights, education, culture and sports, police and homeland security, tax and economic policies, health, and transportation. One theme-employment-was addressed by only two representatives, and the following themes were addressed by only one representative each: racism, the environment, social rights and housing, economy and trade, media and communication, the European Union, law and legislation, and agriculture.
The total number of parliamentary questions analyzed for content was 2,832, of which 1,481 (52.3%) were from UK Muslim representatives, 1,290 (45.5%) from German Muslim representatives, and the remaining 61 (2.2%) were from Israeli Muslim representatives.

Muslim Issues
To test H1 and H2 about the substantive representation of Muslim men and Muslim women, I calculated the number of times Muslim issues appeared in their parliamentary questions.
Within the country. Of the Muslim representatives, 44% asked no parliamentary questions regarding Muslims in their current countries. Since previous studies found that individuals who belong to two or more disadvantaged groups are more invisible in terms of political representation than those who belong to only one disadvantaged group (Hughes 2011;Celis and Erzeel 2017;Hancock 2016), I wondered whether the representatives' invisibility was restricted solely to their descriptive representation or whether it affected their substantive representation as well. Based on my finding, I suggest that invisibility also seems to occur with respect to the substantive component. For example, some Muslim representatives ask very few parliamentary questions about Muslims in their current countries (1%-3%). In contrast, for other Muslim representatives, a large percentage of their parliamentary questions are about such issues (44%-87%). 16 The findings confirm H1 in part because some parliamentary questions by some Muslim representatives dealt more with Muslim issues compared with other issues, while those of other Muslim representatives did not.
To determine whether the variance that exists between individual representatives exists between the three countries, as well, I conducted an ANOVA test that revealed a significant difference (F=99.005, sig=0.00). On average, Muslims in Israel asked more questions about Muslim issues than those in the UK and Germany. In addition, there was no significant difference in this average between representatives from the UK and Germany. Hence, H4 was confirmed. The findings regarding Germany are not consistent with previous studies, which found that immigrant-origin representatives asked more migration-related parliamentary questions and are more likely to engage in the substantive interests of minorities (Wüst 2014;Saalfeld and Bischof 2012). The findings are, however, in line with a previous study on Arab Israeli representatives (mostly Muslims) whose substantive representation refers primarily to Arab citizens in Israel and to a lesser degree, in the Palestinian Authority (Akirav 2014).
The discrepancy between Akirav's study and the current study can be explained by the fact that the former studied all immigrant-origin representatives while I analyzed only Muslim immigrant representatives, a smaller and more coherent group. Second, this study covers 2013-2017, while other studies analyzed earlier periods Wüst 2014). It is therefore possible that the socialization process that took place since changed the focus of substantive representation among Muslim representatives as they are more educated, liberal, and secular, and have fewer children compared with the general Muslim population in Germany. Third, Muslim representatives in Israel are natives and so are part of Israeli society. As such, their Muslim identity is more meaningful compared with their other identities. Furthermore, most of them are well educated 17 , have few children 18 , and are rooted in their identity as Muslims (Ghanem and Mustafa 2011;Jamal 2011). The Muslim identity of Muslim representatives in Israel has not become invisible; on the contrary, it has strengthened and has become interconnected with their Arab identity (Jamal 2011;Akirav 2014). Fourth, procedures regarding parliamentary questions differ in each country: in Germany the procedure is more faction oriented and more publicity oriented than in the UK and Israel, where the decision to use parliamentary questions is entirely up to the individual representative. Fifth, the electoral systems differ as well: Israel has a proportional electoral system without constituencies, while both the UK and Germany have constituencies, although in Germany, the constituencies are in addition to the proportional party list. In all three countries, the party plays a significant role in the candidate selection method and in the daily life of representatives in parliament. Finally, all Muslim representatives in Germany and Israel were from the opposition parties while some of the Muslim representatives in the UK were from the coalition (Conservative) party. Party affiliation is considered to be an explanatory variable in representatives' behavior, as they are obligated to act according to the coalition's policy even when doing so such contradicts their identities.
H5 posited that membership in the coalition or opposition would explain the differences between Muslim representatives and the content of their parliamentary questions. I found a significant difference between coalition and opposition Muslim representatives (F=17.296, sig=0.000), whereby opposition Muslim representatives ask more parliamentary questions about local Muslims than do coalition Muslim representatives. This finding is consistent with previous studies on the use of parliamentary questions by opposition representatives (Jensen et al. 2013;Kolpinskaya 2017).
Outside the country. Of the Muslim representatives, 76% asked no parliamentary questions regarding Muslims worldwide. Two (8%) representatives asked, respectively, 10% and 17% of their parliamentary questions about Muslim issues worldwide, while only 1%-7% of the questions of the rest were about such Muslim issues. These findings did not confirm H1. One explanation stems from the diversity of Muslims worldwide. Muslim countries have different regimes (and most are not democracies) and often struggle among themselves. A second explanation involves the representatives' invisible identity, which may be relevant to a certain degree within the country but irrelevant outside the country. A third explanation refers to the multiple identities, and hence multiple representations, of Muslim representatives that require them to decide which identity to pay more attention to and which to neglect.
There were, however, differences between the countries. Muslim representatives from Germany asked a higher percentage of parliamentary questions about Muslim issues worldwide compared with representatives from the UK and Israel (F=16.839, sig=0.000), confirming H4. Most of the German Muslim representatives were of Turkish nationality and some of their parliamentary questions about Muslims worldwide were about Muslims in Turkey or Muslim refugees from Syria, which is a major topic on the German political agenda. Finally, H5 about opposition/coalition status was not confirmed. Perhaps such a status is relevant to legislators' conduct within the country but does not pertain to issues elsewhere in the world.  (Semyonov et al. 2015;FitzGerald et al. 2018).

Intersectional Representation Index
To visualize the composition of the intersectional representation with respect to Muslim and gender representation, I created an index of Intersectional Representation. As described previously, the classification scheme consisted of twenty categories. To reduce the number of topics, I used a more comprehensive coding based on eight categories.
The index is comprised of eight themes (identities): Muslim women (MW); women in general (WG); Muslims in the country (MIC); Muslims outside the country (MO); the first dominant issue (FI), meaning the theme that received the highest percentage of questions; the second dominant issue (SI), meaning the theme that received the second highest percentage of questions; the third dominant issue (TI), meaning the theme that received the third highest percentage of questions; and all other issues (OI).
The index reveals the proportion of Muslim and gender-related themes followed by the proportion of the three most dominant issues for each representative compared with Muslim and Muslim women's issues. The index is formulated as follows: IR i =MW/n i +WG/n i +MIC/n i +MO/n i +FI/n i +SI/n i +TI/n i +OI/n i Figure 4 demonstrates the intersectional representations of Muslim representatives in the UK, Germany, and Israel according to the eight themes.  As mentioned previously, Purdie-Vaughns and Eibach (2008) argued that people with multiple subordinate group identities are defined as non-prototypical members of their respective identity groups and experience "intersectional invisibility." I wondered whether such representatives "lose" their intersectional identities and concentrate on the one they think is most relevant. Alternately, is it only other representatives and/or the public who are concerned with "intersectional invisibility," whereas the representatives themselves continue to act in all of their identities? Figure 4 shows that intersectional representation exists, but within it each representative decides and acts according to the issues he or she wants to advance. For example, Israeli Muslim representatives concentrate on their Muslim identity, while UK Muslim representatives neglect their Muslim identity in their use of parliamentary questions. German Muslim representatives pose more parliamentary questions regarding their Muslim identity compared with UK Muslim representatives, but do so significantly less than do Israeli Muslim representatives.

Conclusions
The gap in political representation between represented and non-represented identities is defined as a "double disadvantage." It leads to greater invisibility among representatives who belong to several disadvantaged groups and have intersectional identities, compared with those who belong to only one disadvantaged group.
The study showed that all Muslim representatives address a variety of issues, represent more identities, and also address Muslim issues or women's issues. However, the three countries differ significantly. While Muslim representatives in Israel are more coherent regarding their Muslim identity, Muslim representatives in Germany and in the UK address religious issues but have other identities to represent as well. Furthermore, Muslim women's issues are generally invisible in the UK and Germany, whereas in Israel they are addressed more often. Muslim representatives in Western countries have intersectional identities and, in some cases, their Muslim or Muslim female identity becomes invisible. Representatives face a never-ending intersectional struggle when they have intersectional identities to address. Hence, as realistic political actors, they must decide which identity to represent and which to ignore.
The findings have several implications for the representation literature. Representation theory focuses mostly on gender and race/ethnicity, but less often on religion. Hence, the study expands our understanding of the role that religion plays when it is included in intersectional identities. Future studies should include religion as an inherent part of representatives' identities. Another implication is the inclusion of Israel in the representation literature, which is largely based on the Anglo Saxon and West European experience. In addition to expanding this literature, adding Israel also allows us to compare the actions of indigenous or native minorities with those who are immigrants. Indeed, many studies focus on those with a migration background only. Future studies should expand these comparisons and include more democracies with Muslim populations from Europe and from other continents, such as Canada and the US or Australia where most Muslims are not immigrants. Future studies should investigate whether the racial/ethnic status of representatives influences their intersectional representation, because religion can involve issues of both race and gender.
Finally, the extensive analysis of nearly 3,000 parliamentary questions expands our understanding of substantive political representation. In parliamentary systems parliamentary questions are the most significant tool that individual representatives can use to oversee the government with minimum control by the party. Hence, future studies should concentrate on analyzing the content of parliamentary questions from larger numbers of representatives, and regarding the terms analyzed and the countries examined.