Europe’s Growing Muslim Population
Appendix A: Methodology
The estimates and projections in this report build upon and update data from the April 2015 Pew Research Center report, “The Future of World Religions: Population Growth Projections, 2010-2050.” Soon after the release of that report, a large influx of migrants entered Europe seeking refugee status. This report includes estimates of how Europe’s Muslim population changed from mid-2010 to mid-2016, with particular attention to change caused by migration. It also includes projections of how the size of Europe’s Muslim population may change in the future under different migration scenarios.
This study takes advantage of more than 2,500 data sources gathered for previous projections, including censuses, demographic surveys, general population surveys and other studies. Additionally, this study draws on new data, including asylum-seeker data through mid-2016 from Eurostat (Europe’s statistical agency), as well as new survey and other government data.
Baseline (2010) estimates of populations by religion
The earliest population figures in this report are for the year 2010. Several Pew Research Center reports estimated the size of religious populations, including Muslims, in Europe in the year 2010 (“The Future of the Global Muslim Population,” released in 2011; “The Global Religious Landscape,” released in 2012; and the aforementioned “The Future of World Religions” report, released in 2015). Each report builds on our prior reports, adjusting estimates based on new data. While previous religious demography reports classified 50 countries, including Russia, as part of Europe, this report focuses on a narrower set of 30 countries: the 28 member nations of the European Union (as of mid-2016), as well as Norway and Switzerland.7 The reasons for the narrower geographic focus of this report are both substantive and pragmatic. These 30 countries have received the bulk of Europe’s asylum seekers in recent years and they are the countries for which data on asylum-seeking patterns are available from Eurostat.
The gold standard for measuring religious identity in this report is a census or survey question that asks, “What is your religion, if any?” The aim in this report is to measure identity sociologically rather than theologically. Individuals who self-identify as Muslim are classified as such, regardless of their level of adherence to what might be considered orthodox belief and practice.
Among the 30 countries covered in this report, about half of these nations directly measure religious identity in a traditional census or census substitute (such as a large-sample household survey). Census data are ideal for measuring the size and characteristics of minority populations. However, in the remaining countries it was necessary to rely on sources that lack the statistical power of a census. In some countries, general population surveys and demographic surveys provide sufficient detail on the size and demographic characteristics of Muslim populations. However, typical general population surveys of 1,000 to 3,000 respondents may under-sample Muslims, particularly in countries with substantial first-generation immigrant populations, who may not be fluent in the country’s dominant languages and who may be difficult to capture in standard sampling frames. Estimates of the size of Muslim populations are based on an assessment of all available data, including census and survey data, population registers, immigration data and other reports and sources. Primary sources used for each country are listed in Appendix B.
Special considerations in France and Germany
The sensitivity of measuring religious identity varies across European countries. While many countries do collect data on Muslims and other religious groups in a census, in other countries, such as France, governments restrict the collection of religion data on the census and other government surveys. A 1978 French law imposes limitations on the collection of data pertaining to race, ethnicity and religious opinions unless the subject gives express consent. While religion has not been measured on a nationwide government census in France since 1872, it is nonetheless still possible to measure religious identity and practice in France.
Two French surveys are particularly important for this report. Our baseline estimate of the size of France’s 2010 Muslim population is based primarily on data from the 2008 “Trajectories and Origins” survey of more than 20,000 respondents in metropolitan France (which includes oversamples of first- and second-generation immigrants) sponsored by the French Institute for Demographic Studies (INED) and the National Institute of Statistics and Economic Studies (INSEE). Our projection to 2016 was validated against a 2016 survey of 15,459 respondents sponsored by the Institut Montaigne and carried out by the French Institute of Public Opinion (Ifop).8
Since Germany’s 1990 reunification, no German census was conducted at all until 2011, when the country was required to carry out a census as a member of the EU. Prior to 2011, population estimates had been based on reports of births, deaths and moves (in, out or within Germany), which people are required to make to their local governments. Individuals who left Germany without an expectation of returning were not motivated to deregister as residents, and were over-counted, artificially exaggerating the “healthy-migrant effect” – the puzzlingly high number of migrants living to be 110. While aggregations of municipal registries were adjusted to account for some failure to deregister, such adjustments were not sufficient, and the 2011 census revealed that Germany had overestimated its population by 1.5 million people. Most of these missing people were migrants who had left Germany without deregistering. The 2011 census was an improvement, but it did not gather accurate estimates of religious group sizes. Due to a problematic questionnaire design, religious affiliation was measured with a two-step question, which was partly optional, and the results were vague and systematically undercounted non-Christians. Our estimates of Germany’s Muslim population in this report build upon new estimates from the German Federal Office for Migration and Refugees.9
The role of projections in this report
The 2016 estimates and 2050 projections in this report are based on projections from 2010 baseline data. The projections rely on input data about the characteristics of Muslims and non-Muslims, including age and sex composition, fertility rates, religious switching patterns, and migration patterns. The next section describes our projection methods. Subsequent sections provide detail on the input data used in the projections.
The projection approach: Explaining multistate cohort-component projections
The technical calculations for the projections in this report were made by Marcin Stonawski, the Religion-Education-Demography project leader at the International Institute for Applied Systems Analysis (IIASA), in consultation with Michaela Potančoková (a research scholar at IIASA) and Pew Research Center researchers, using an advanced variation of the standard demographic method of making population projections. The standard approach is called the cohort-component method, and it takes the age and sex structure of a population into account when projecting the population forward in time. This has the advantage of recognizing that an initial baseline population can be relatively “young,” with a high proportion of people in younger age groups or relatively “old,” with a high proportion of older people.
Cohorts are groups of people that had an experience in a particular time. A birth cohort, the type of cohort referenced in this context, comprises people born in a certain period. Birth cohorts can also be described as males or females who have reached a certain age in a particular year. For example, the cohorts of females ages 15 to 19 in the year 2000 and males ages 15 to 19 in the year 2000 shared the experience of being born between 1981 and 1985.
Components are the three ways in which populations grow or shrink: new entrants via births, exits via deaths and net changes from migration. Each cohort of the population is projected into the future by adding likely gains – births and people moving into the country (immigrants) – and subtracting likely losses – deaths and people moving out (emigrants) – year-by-year. The very youngest cohorts, those ages 0 to 4, are created by applying age-specific fertility rates to each female cohort in the childbearing years (ages 15 to 49).
The cohort-component method has been in existence for more than a century. First suggested by English economist Edwin Cannan in 1895, then further improved by demographers in the 1930s and ’40s, it has been widely adopted since World War II. It is used by the United Nations Population Division, the U.S. Census Bureau, other national statistical offices, and numerous academic and research institutions.
The advanced variant of this approach, multistate cohort-component projection, became viable starting in the 1970s thanks to the availability of mainframe computers and work by the American geographer Andrei Rogers, among others. The multistate approach permits simultaneous projection of multiple religious groups, taking into account variation by religion in age, sex, childbearing patterns and propensity and direction of migration. This approach also enables modeling of religious switching as a transition between religious “states.”
Projection inputs for each country – including differential data by religion on fertility, age structure, migration and, where available, switching rates – were used for multistate cohort-component projections going out to the year 2050. Country-level 2010 population data, as well as fertility and mortality trajectories, are based on the 2015 revision of the United Nations’ World Population Prospects.
All projection models assume that Muslims and non-Muslims within each country will see their fertility levels slowly converge toward identical fertility rates by 2110 – a century from the baseline year of the projections. The assumption that fertility differences gradually will diminish within countries is based on evidence that when people live in the same economic and social milieu, their fertility patterns tend to become increasingly similar over time. Studies have shown, for example, that the offspring of immigrants to the United States and Europe tend to adopt the fertility patterns of the general population in the countries where they live within a few generations. The adoption of a relatively conservative 100-year timeframe for within-country convergence reflects the fact that geographic clustering, differing education levels and other factors may perpetuate distinctive childbearing patterns among some religious groups.10 At each step of the main projection scenario, fertility for the total population of a country follows the UN medium variant assumptions from the 2015 revision.
All projection models extend current religious switching rates in the 22 countries with available data. Those rates are used to calculate the flow of people in certain age and sex groups who move between Muslim and non-Muslim categories at each five-year interval of the projections. In countries for which switching data are available, researchers generated recent rates of switching. The main projection model assumes that emerging cohorts will switch from their childhood status at the same rate observed in recent survey data.
Based on projection assumptions from the United Nations Population Division, this study projects that life expectancy at birth will gradually increase in all countries. There is no high, medium or low assumption because each country, regardless of its current economic condition, is assumed to be moving toward better living standards and, therefore, longer life expectancy at birth. Following the 2015 revision of UN projection assumptions, gender-specific differences in mortality are introduced based on the UN assumptions of life expectancy by sex.
Disclaimers about projections
Some cautionary words are in order. Population projections are estimates built on current population data and assumptions about demographic trends. The future of the Muslim population in Europe will be influenced by economic and political circumstances in Europe that affect the feasibility and desirability of immigration, as well as circumstances outside Europe, including political upheavals and armed conflicts that could produce migration surges. The future of Muslim and non-Muslim populations also may be influenced by scientific discoveries, environmental challenges and other changes that could shift demographic trends in unforeseen ways.
Estimating migration is difficult. Projecting it is even harder. The mass movement of people can be the result of several different push and pull factors, including economics, politics and conflict. Predictions are hard to make and when unexpected migration-related events occur, as they have in and around Europe during the past few years, sudden change can immediately alter the number, origins and destinations of future migrants.
These migration uncertainties have been part of academic discussions for years. Most recently they were studied by Oxford University’s Global Migration Futures project. This study brought together migration experts to map migration factors that could have the greatest impact on future migration flows into and out of European countries. They also categorized these factors by their level of uncertainty. Several of the factors involved changes in Europe’s economy, the political integration (or potential disintegration) of the European Union and conflict in surrounding regions.
The projections in this report are not meant to forecast the future, but instead present estimates for the religious composition of Europe under three migration scenarios to convey a range of “what if” outcomes.
The medium and high projection scenarios in this report assume that in the future, the countries sending migrants to Europe will remain the same as they have been in recent years. Of course, this will not be exactly correct. What will be consequential for the size of the Muslim population in Europe will be whether countries sending migrants in the future continue to be countries with large Muslim populations. For example, if Turkey replaces Syria as a top country of origin in the future, this would continue to boost Muslim numbers, whereas an influx of migrants from Canada or China probably would not have the same effect on Muslim numbers. And while Africa’s large Muslim populations tend to be geographically closer to Europe than its large Christian populations, a change in the religious mix of Africans entering Europe could also have a large impact.
Input data for population projections
The demographic projections in this report use data on age and sex composition, fertility, mortality, religious switching and migration. This section describes how these data were gathered and standardized for use in the projections.
Estimating age and sex structures
Religious affiliation varies by age. In this section, the phrase “age structure” is used as shorthand to refer to the religious composition of age-sex groups. In order to calculate the median ages of religious groups and carry out population projections, researchers assembled age structures for Muslims and non-Muslims in every country. Data on age structures were collected in 20 age categories (measured in five-year increments with a top value of 95 and above) for males and females (e.g., males ages 15 to 19), resulting in a total of 40 categories.
Researchers constructed initial age structures by analyzing survey datasets, census datasets and tables published by census agencies. While censuses usually enumerate religion for the entire population, including children, general population surveys do not usually include interviews with children. Since age structures require religious affiliation data for children, children were assigned religious affiliations when necessary based on the best methods available, including estimating the religious affiliation of children based on the fertility patterns and religious affiliation of women of childbearing age, as well as information about the religious affiliation of the youngest respondents measured in a survey.
In many countries, there are substantial differences in the number of children born to Muslim and non-Muslim women. Furthermore, groups often vary in the share of women in their population who are of childbearing age, and women in some groups may, on average, begin having children at younger or older ages than do women in other groups. These differences in childbearing patterns, age structure and fertility timing combine to produce differences in the rates at which babies are born to Muslims and non-Muslims.
Fertility data were gathered from censuses and surveys, and fertility rates were estimated via direct and indirect measures. Some censuses and surveys directly measure recent births or the number of children a woman has ever borne by the time of the survey. In other cases, fertility data were gathered indirectly, for example, by using data on the age of a mother’s children to estimate her past birth patterns. These various sources of fertility data were used to estimate age-specific and total fertility rates for Muslims and non-Muslims in each country.
Fertility rates were estimated for the 2010 to 2015 period and projected forward to subsequent time periods with differences between Muslims and non-Muslims slowly converging. In the absence of data on the fertility rates of new Muslim migrants, they are assumed to have the prevailing fertility rates of other Muslims in their destination countries.
The estimates of differential fertility for Muslims and non-Muslims in the 30 countries of this report are the same as those estimated for the 2015 “Future of World Religions” report.
Each country’s projected mortality patterns are based on UN mortality tables for the country. Little research has been conducted on cross-national differences in mortality and life expectancy across religious groups. In the absence of better data, the same mortality patterns within each country are assumed for Muslims and non-Muslims. Muslims in many European nations have less education than non-Muslims, which could be associated with lower life expectancy, but a large share of many European Muslim populations are first-generation migrants, and being a first-generation migrant is sometimes associated with health advantages and longer life expectancy relative to native-born residents of destination countries (for example, Hispanic migrants to the United States have advantages compared with those born in the U.S.).
Estimating religious switching
Studies of religious switching indicate that this phenomenon is often concentrated in young adult years, roughly between the ages of 15 and 29. Change in religious affiliation may occur as young adults move away from their parents and forge their own identity or partner with someone of a different affiliation status. While some religious switching may take place at other ages, switching is modeled as a life course phenomenon in which some young adults change their religious affiliation status. There may be some time periods during which people of all ages are prone to religious switching, such as when political circumstances in a country encourage or discourage religious identity or lack of religious identity. Our models do not attempt to include such period effects.
Sources of switching data
The typical procedure for measuring religious switching is to compare the religion in which a person grew up with their current religion (when the person is an adult). The best sources of data on religious switching are nationally representative surveys that ask adults about their current religion and the religion in which they were raised. These surveys typically have sample sizes between 1,000 and 3,000 respondents. Unfortunately, while censuses and large-scale demographic surveys often measure current religious affiliation, they generally do not measure religious origins, so they cannot be used to directly measure religious switching. (Censuses in Northern Ireland and Scotland are exceptions.)
In 22 countries (Austria, Belgium, Croatia, Czech Republic, Denmark, Finland, France, Germany, Hungary, Ireland, Italy, Latvia, Netherlands, Norway, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, Switzerland and the United Kingdom), adequate survey data are available on both the religious upbringing of survey respondents and on current adult religious identity.11 Data on patterns of switching from non-Muslim religions to Islam come from the International Social Survey Program. In France, data on patterns of switching out of Islam among those raised Muslim come from the Trajectories and Origins survey, which found that approximately 10% of those raised Muslim later switched to no religious affiliation or to some other religion. Because other European countries lack sufficient data to measure patterns of switching out of Islam, the French pattern is borrowed for other Western European countries. Thus, in Western Europe, projections model that roughly 10% of rising cohorts of children raised in Islam will switch out of the faith as adults.
Since men and women often follow different switching patterns, researchers calculated rates of switching separately for men and women based on the experiences of adults ages 18 to 54 at the time of the survey. Researchers assume that the experience of young respondents is the best source of information about likely switching patterns for emerging generations, so the experiences of older respondents (those ages 55 and above) are excluded from the analysis. The analysis was initially restricted to the switching experience of 30- to 54-year-olds; while this restriction allowed the focus to be on respondents who have recently completed their young adult years, it left less-than-optimal sample sizes. Including the full range of adults ages 18 to 54 in the sample increased sample sizes and did not appear to compromise the reliability of the switching rates.
The estimates of religious switching for Muslims and non-Muslims in the 30 countries of this report are the same as those estimated for the 2015 “Future of World Religions” report.
Estimating and projecting the size and religious composition of regular migrants
To model the impact of migration on future religious change, the population projections in this report required an estimate of the religious composition of recent migrant flows between countries. That is, how many migrants moving from country X to country Y are Muslim? How many have some other religious identity? Data on the size and religious breakdown of migrant flows were pieced together in two steps. The first step was to estimate how many people move to and from every country in the world. Second, the religious composition of migrants moving between countries was estimated.
Generally speaking, there is much better information on migrant “stocks” (how many foreign-born people reside in each country, and where they were born) than there is on migrant “flows” (how many people move between countries each year). The limited flow data that are available may not capture all modes of travel or all kinds of international migrants, and it can be difficult to distinguish short-term travel from long-term migration. Since data on migration flows are incomplete, data on migrant stocks for 2010 and 2015, estimated by the United Nations, were used to estimate migration flows for both males and females.12 Demographer Guy Abel developed an innovative technique to estimate migration flows between countries using this stock data.13 Using empirical data and observed regularities in the age patterns of migration flows, researchers were able to disaggregate each estimated total flow into subtotals by five-year age groups. The bilateral flows estimated based on the UN migrant stock data include asylum-seeking and regular, non-asylum-seeking migration. Although the UN data include refugee stock estimates from the Office of the UN High Commissioner for Refugees, the latest UN migrant stock data (released in December 2015) was prepared too early to capture the large volume of asylum seeker flows that arrived late in the 2010 to 2015 interval. Since Eurostat provides updated measures of asylum seeker flows, Eurostat is the source of asylum seeker data in this report and UN data were manually adjusted to reduce flow estimates (such as the flow from Syria to Sweden) likely to have been largely composed of asylum seekers. Thus, adjusted data from Guy J. Abel, based on UN stock estimates, became the source for estimating regular migration; refugee flows, adjusted based on rejection rates, were estimated separately based on Eurostat asylum seeker data.
Another step was to identify the religious breakdown of migrants. The religious composition of migrants is not always the same as the religious composition of the general population in their country of origin. In many cases, members of some religious groups are more likely than others to leave a country, and they are also more likely to choose certain destination countries. Religious minorities, in particular, may be disproportionately likely to migrate to a country in which their religion is in the majority. The religious breakdown for the movement of migrants is drawn from Pew Research Center’s Global Religion and Migration Database – which has estimates of the religious breakdown of migrant populations based on global census and survey information.
Using all of this information, researchers calculated migration rates to and from countries by age, sex and religion. Using migration rates instead of population counts allows for a more dynamic model of future migration. As countries increase or decrease in size and their religious composition changes, the migration rates will produce corresponding changes in the size and religious composition of migrant flows.
Estimating asylum seekers
Estimates of asylum seekers and refugees are based on Pew Research Center’s analysis of Eurostat data on applications for asylum. Eurostat is Europe’s statistical agency, a central repository for high-quality population and other data. Eurostat collects data from 28 EU countries and four European Free Trade Association countries (Iceland, Liechtenstein, Norway and Switzerland) on the number of applications for asylum submitted, withdrawn, accepted and rejected, and it makes these counts available to the public. These publicly available counts of asylum seekers by application status are not linked together, however, so it is not possible for analysts to track individuals or cohorts of applicants through the process.
The estimates of the number of asylum seekers in this report are based on first-time applications for asylum, adjusted for withdrawals (read more about withdrawn applications below). First-time applicants are those who have never previously applied for protection within a country, as opposed to applicants who are appealing a previous decision.
Applications are made at the individual level, rather than at the household level. Thus, every person, regardless of age or family relationship, is required to submit an application for asylum. Applications are not the same as arrivals. It can take several months for newly arrived asylum seekers to make formal applications for asylum and become included in asylum applicant statistics.
Adjusting for withdrawals
Importantly, applicants are not always applying for protection in Europe for the first time, and might file applications in more than one country. Countries in the EU, as well as Norway and Switzerland, have agreed to the Dublin Regulation, which states that asylum seekers are to apply for refugee status in the first European country they enter. The Dublin Regulation is intended to prevent multiple applications and reduce the number of asylum seekers moving from country to country, but it was partially suspended in 2015. Consequently, many asylum-seekers traveled through Greece, making their way north and west to Germany and other destinations using a route through several Balkan countries and through EU countries, such as Hungary and Austria. Some asylum seekers applied for refugee status in transit countries, sometimes unwillingly, before making it to their desired destination. Many asylum seekers who applied in Hungary and Bulgaria, for example, later had their applications withdrawn.
To estimate the number of asylum seekers, rates of withdrawn applications by nationality and country of application pair were calculated from Eurostat’s withdrawn application data and subtracted from the total number of first-time (within country) applications for asylum. Applications are withdrawn either intentionally by applicants or automatically whenever an applicant fails to complete the next step of the application process. These withdrawn applications are not included in this report’s counts of asylum seekers because it is likely that asylum seekers who withdrew their applications either left Europe or applied in another country. Removing withdrawn applications avoids double counting of asylum seekers.
Estimating refugee counts
Refugee counts are estimated based on Eurostat’s quarterly decision data on approved first-time asylum applications.14 Rates based on acceptances between origin country X and destination country Y were applied to every nationality and country of application pair. Applicants who have been accepted, or whose applications are pending and are likely to be accepted based on applicants’ nationality and country of application are counted as refugees for Pew Research Center’s estimates. Acceptance rates were based on all positive decisions. No distinctions are made based on the kind or length of stay or humanitarian protection status granted to the applicant.
Estimating the religious composition of asylum seekers
Though religious identity is frequently relevant and provided in asylum applications, applicant religious affiliation is not collected for its own sake and the publicly available data from Eurostat do not include information on applicant religion. This report’s estimates of religious affiliation of asylum seekers and refugees are therefore based on religious demographic data from other sources.
In some cases, the religious composition of migrants are estimated based on the religious composition of their country of origin, and the models in this report assume that migrants are a random selection of people from a country with regard to religion. When available, data on the religious composition of prior migration flows by nationality and country of application pairs is utilized from Pew Research Center’s Global Religion and Migration Database. In the latter case, models assume that recent migrants’ religious composition is more similar to past migrants’ from that country of origin than the country of origin’s overall religious makeup.
Return migration for refugees
This report does not explicitly model the return migration of refugees after they have settled in Europe based on the assumption (and historical precedent) that even if conditions improve in the country they left, most will continue to remain.15 This assumption is based on the experience of waves of Turks and North Africans who came to Europe during times of labor shortage in the 1950s and 1960s and continued to remain in Europe even when European economies had fewer jobs for them during the 1970s. As noted by migration scholar Philip Martin, “There is nothing more permanent than temporary workers.” Even though the motivation among refugees for coming to Europe may differ than guest workers in earlier decades, refugee families are not expected to uproot themselves in large numbers to return to their home countries. However, if countries enforce the temporary legal status of some refugee decisions, then forcible repatriation, like voluntary returns, could curtail the estimates in this report. While the decision not to model return migration for refugees could have the effect of inflating future Muslim population projections to some degree, this bias may be counterbalanced by the conservative treatment of asylum seekers who are not expected to initially gain legal status to remain in Europe (discussed below).
Estimating legal limbo populations
Finally, recent rejection rates recorded by Eurostat were used to estimate the number of people in Europe who have applied for asylum but have not been and are unlikely to be granted legal refugee status. This is the population of asylum seekers in legal limbo, for whom the future is uncertain. Many applicants who are initially rejected will appeal their decisions, and some may eventually be approved to live and work in Europe as protected refugees. Others will leave, voluntarily or through forced deportation. Still others will obtain another legal status, for example through student or worker visa programs. The population of Muslims estimated to be in legal limbo in 2016 is not counted in the estimates or projections in this report. Some will no doubt stay in Europe, legally or illegally. Not counting this population makes our projections conservative. This conservative bias may help balance the assumption that refugees granted stay in Europe will not return (see preceding section).
Using projections for 2016 estimates and beyond
Generally speaking, most of the 2016 Muslim estimates in this report are the result of projections from 2010 that account for Muslim and non-Muslim differences in age structure, fertility, religious switching and regular migration.16 Additionally, the 2016 figures include estimates of Muslim refugees who came to Europe between mid-2010 and mid-2016. In some cases, 2015 or 2016 government data for Muslim populations were used rather than projections from 2010. For example, a report from the German Federal Office for Migration and Refugees on the year-end 2015 size of the Muslim population in Germany was the basis for our midyear 2016 estimate in Germany (our estimate also incorporates growth in the population from the ongoing flow of refugees in the first half of 2016).
Projection scenarios for 2016 to 2050
To highlight how different migration patterns may change the landscape of Europe’s Muslim population, the only factor that varies in our 2016 to 2050 projection scenarios is migration. The zero migration scenario assumes there will be no migration to or from any of the 30 countries after midyear 2016. Recent history since mid-2016 suggests this scenario will not be a realistic model of migration outcomes. However, the benefit of this scenario is to highlight how factors other than migration would be expected to change Muslim population numbers in Europe. The medium scenario models change that would be expected if only regular migration (i.e., no asylum seeker flows) continued, following the estimated patterns of country origin and destination flow from the 2010 to 2015 period that were used to make 2010 to 2016 projections. The high migration scenario adds an ongoing flow of refugee levels seen between the beginning of 2014 and mid-2016 (i.e., the continuation of what some have described as a “refugee crisis”) in addition to regular migrants. For example, in the high scenario, Germany is projected to receive 200,000 Muslim and 30,000 non-Muslim refugees annually because these are the annual average of the estimated numbers of refugees Germany received from January 2014 through June 2016.
Sex balancing in high migration scenario
Recent flows of asylum seekers have been dominated by young men (roughly 70% to 75% of asylum seekers have been male). Under certain conditions, refugees who obtain humanitarian protection status are entitled to family reunification. If current refugees are able to bring family members to Europe, this could increase the female share of future refugee flows as spouses, children and other relatives join existing refugees.
Information on the family composition (including marital status) of existing refugees is not available, so family reunification has to be modeled hypothetically rather than on the particular family characteristics of refugees. The approach used in these projections assumes many refugees will seek to bring family members and/or spouses who share their cultural background (i.e., from their home country) rather than building new family structures in the host society.
Gender balance ratio targets were created for each of the 30 countries based on 2011 population census data about the stock of the foreign-born population aggregated by Eurostat by nationality and country of application pairs. This assumes that over time, family reunification for refugees could permit the gender balance of refugees in a country to approach the gender balance of first-generation immigrants in the country. Typically, the 2011 stocks of first-generation immigrants in Europe were roughly gender balanced, with women generally making up 44% to 55% of this population, depending on the nationality and country of application.
For refugees projected to arrive until mid-2025, the flow of refugees in the 15-to-49 age categories is assumed to have a gender composition that will complement the stock of refugees who arrived since 2010 so that by 2025, the overall stock of refugees (since 2010) of reproductive age in each country will be roughly gender-balanced. After 2025, the high migration scenario assumes that further flows of refugees will be gender-balanced.
- For the 50-country Europe region in “The Future of World Religions,” the Muslim population in 2010 was estimated to be 43.5 million (when Cyprus – which is part of the European Union but is not included in Europe in that report – is added in, the total rises to 43.7 million). Of the remaining 20 countries excluded from this report, the most consequential difference is the absence of Russia, which had an estimated 14.3 million Muslims in 2010.
When Muslim estimates used in “The Future of World Religions” report are aggregated for the 30 countries in this report, the 2010 total is 21.2 million (and 22.2 million for the remaining 20 countries, including Russia). In this report, the 2010 estimate for the size of the Muslim population in the EU, Norway and Switzerland is 19.5 million. The country estimate that is most different between reports is Germany. “The Future of World Religions” had a higher estimate of Germany’s Muslim population in 2010 (4.8 million) because it relied on estimates of Germany’s overall population size that analysis of the 2011 census found to be exaggerated (see discussion of Germany in the next section). Based on analysis of new government data, this report estimates Germany’s Muslim population in 2010 to have been 3.3 million. ↩
- The report based on the Institut Montaigne survey does not estimate the total size of the Muslim population, including children, but data from the survey confirm that Muslim shares are much higher among the youngest respondents than among older French respondents. If the incidence of Muslim identity among those younger than 15 years old in France is similar to incidence among respondents ages 15 to 17 in the survey, then when that incidence rate and the incidence rate for other cohorts in the survey are multiplied by the size of each cohort in 2016, the estimate for the total population is an overall incidence rate slightly below 8%. This report’s projection based on the Trajectories and Origins baseline is that the Muslim incidence rate reached 8.8% in 2016. Working with the same baseline survey data, French demographer Michèle Tribalat finds estimates of France’s Muslim population based on the 2016 Ifop/Montaigne data to be conservative. ↩
- “The Future of World Religions” report had a higher estimates of Germany’s Muslim population in 2010 (4.8 million) because it relied on estimates of Germany’s overall population size that analysis of 2011 census found to be exaggerated (see discussion of Germany in the next section). Based on analysis of new government data, this report estimates Germany’s Muslim population in 2010 to have been 3.3 million. ↩
- In European countries with longstanding Muslim populations that have long co-resided with non-Muslims, Muslims still tend to have modestly more children than non-Muslims, even after controlling for socioeconomic differences. See Stonawski, Marcin and Michaela Potančoková and Vegard Skirbekk. 2016. “Fertility Patterns of Native and Migrant Muslims in Europe.” Population, Space and Place. ↩
- Switching data were not available for Bulgaria, Cyprus, Estonia, Greece, Lithuania, Luxembourg, Malta and Romania. ↩
- This approach is described in Abel, Guy J. Forthcoming. “Estimates of Global Bilateral Migration Flows by Gender between 1960 and 2015.” International Migration Review. ↩
- See Abel, Guy J. 2013. “Estimating global migration flow tables using place of birth data.” Demographic Research. Also see Abel, Guy J. and Nikola Sander. 2014. “Quantifying Global International Migration Flows.” Science. ↩
- Appeal rates are more difficult to generate over a series of years. Application acceptance rates used in this study do not take into account asylum seekers who may later be granted permission to stay in Europe after an appeal. Consequently, acceptance rates used in this study are conservative. ↩
- General migration out of European countries, based on emigration estimates derived from Guy Abel’s analysis of United Nations migration stock data, is part of all projection scenarios except the no migration scenario. These emigration estimates assume that patterns from the 2010-2015 period will continue. ↩
- Projections of religious projections from Pew Research Center typically use five-year projection increments. In order to report on Muslim population estimates after the large influx of asylum seekers in 2015 and 2016, midyear 2016 estimates are based on refugee estimates, projections from 2010 to 2015 as well as roughly one-fifth of the change projected for the 2015 to 2020 period. ↩