Innumeracy in Turkey: Misperceptions of an Emerging Immigrant Population

Introduction

The World Bank estimates that after 2007, the nation of Turkey transitioned from experiencing levels of net emigration to net immigration.¹ Turkey thus follows Italy, Spain, Portugal, and Finland, which underwent similar transitions during the 1980s and 1990s.² In a recent issue of Insight Turkey, Franck Düvell describes these changes as ushering in a paradigm shift of “enormous social, economic and political relevance,”³ which will require an increased research focus. Indeed, the transition from receiving more immigrants than it sends makes Turkey an interesting research context, not only locally, but for the study of immigration in general. I heed Düvell’s call for a greater focus on Turkey as a recipient of immigrants by analyzing how accurately Turkish respondents perceive their immigrant population and whether misperceptions are connected to anti-immigrant attitudes.

While discussing Turkey’s immigration shift, Düvell notes that “Academics were the first to highlight these developments… However, the Turkish public is probably less aware of this shift…” (emphasis added).⁴ The italicized statement is reasonable, but it contradicts a recent line of scholarship within the intergroup attitudes literature. There is a well-documented tendency, known as immigrant population innumeracy, wherein most ordinary citizens wildly over-estimate immigrant population sizes.⁵ However, this phenomenon has only been considered in detail in the U.S. and Europe, which have been contexts of net immigration for several decades. 

---

Regardless of the context, immigrant populations are larger in the minds of citizens than in reality

---

With this in mind, Turkey is a useful context for analysis. Such a project would mark the first assessment of innumeracy in the Middle East and the first in a majority Muslim country. It would also be the first detailed analysis in a nation that sits on the cusp of an immigration transition. This unique characteristic may help us to understand various phenomena related to immigration in general. For one, when and how do citizens develop their perceptions and beliefs about immigrants? Focusing on the Turkish context at the beginning of its paradigm shift will elucidate this question. My goal is to examine innumeracy among Turkish citizens, and in doing so, inform the wider literature on misperceptions about immigrants.

Do Turkish respondents display innumeracy like those in the U.S. and Europe? If Düvell’s assumption is correct, Turks may actually underestimate their immigrant presence. Further, if misperceptions exist, do they result in any consequences in terms of attitudes about immigrants? Using data from the 2013 Transatlantic Trends Survey, the current study considers Turkish innumeracy levels in detail, focusing on (1) how much exists, (2) where it is most common, (3) among whom it is most likely to be expressed, and (4) whether it is associated with anti-immigrant attitudes.

Immigrant Population Innumeracy

When researchers ask how many immigrants they think are living in their country, respondents usually offer an inflated estimate. In the most recent data from Ipsos MORI,⁶ respondents in all of the 14 countries considered overestimated the immigrant population size on average. For example, in the U.S., where immigrants represent roughly 13 percent of the country, the average guess was 32 percent.⁷ In other words, the typical American thinks the immigrant population is about three times its actual size. With an average level of incorrectness of 19 percentage points, the U.S. ranks behind only Italy, where respondents were wrong by 23 percentage points.⁸ In Australia and Sweden, which displayed the greatest accuracy, respondents still believed that the immigrant population was seven percentage points larger than the reality. Thus, regardless of the context, immigrant populations are larger in the minds of citizens than in reality.⁹ 

Nearly all of our knowledge about immigrant population innumeracy comes from European and American samples. Thus, to advance this research, it is important to examine whether the same patterns exist in other parts of the world. A unique exception is found in the same Ipsos MORI survey, which indicated that respondents in Japan and South Korea also inflate immigrant population size by about 8 percentage points on average.¹⁰ Thus, while there is evidence that innumeracy exists outside of the West, more research is needed to understand in what other parts of the world these misperceptions flourish, why such inaccurate knowledge exists, and whether there are consequences to widespread ignorance. 

Why Innumeracy?

Interest in innumeracy among social scientists developed from its connection to other variables. After all, if innumeracy were unassociated with demographic, social or attitudinal factors, then it could be attributed to simple random ignorance and would be of little consequence.¹¹ However, fully random innumeracy has never been the case in the literature. Rather, overestimation is frequently linked to various demographic variables. In both the U.S. and throughout Europe, women generally inflate population sizes more than men, younger people tend to overestimate more than older generations, and the more highly educated tend to offer more realistic size estimates.¹² Because of such associations, researchers generally accept that innumeracy represents more than simple random ignorance.

Beyond demographic characteristics, scholars have also theorized psychological sources of innumeracy related to the widespread use of heuristic decision-making.¹³ These mental shortcuts allow individuals to quickly make judgments and come to decisions with minimal effort. In particular, the cognitive availability heuristic, introduced by Tversky and Khanneman,¹⁴ describes cases in which individuals use familiar examples in their minds as evidence to answer questions of fact. For example, if one has many acquaintances that have been through a divorce, he or she is more likely to overestimate the rate of divorce. Transferring this logic to questions of immigrant population size, those with more interpersonal contact with immigrants (friends, neighbors, coworkers, etc.) will have a larger perception of the immigrant population. Previous research in Europe confirms this pattern.¹⁵

Of paramount interest to researchers has been innumeracy’s connection to negative attitudes about the groups being estimated. There is concern that warped perceptions about the reality of immigration will be used to justify or generate anti-immigrant actions and policy preferences. Individuals may conclude that immigrants are too numerous or pose a threat to citizens based on incorrect information. Consistently across studies, larger size estimates are associated with more negative attitudes toward immigrants and support for hypothetical policies designed to curtail immigration or limit immigrants’ rights.¹⁶

A view from one of the Syrian refugee camps, built by AFAD (Turkish Prime Ministry Disaster and Emergency Management Authority) in Suruç, Turkey. / AFAD Press Office

Given the lack of research on innumeracy in the Turkish context, it is unknown whether these same patterns will hold. If innumeracy exists among ordinary Turks, is it simple random ignorance or is it associated with other factors? Further, is innumeracy connected to negative attitudes? I consider these questions below after discussing the Turkish immigration context. 

Immigration and Innumeracy in Turkey

For several decades, Turkey has been characterized as a country of emigration. Turkish citizens have settled as labor migrants throughout the world, particularly in Germany, other parts of the European Union, and the Persian Gulf states.¹⁷Subsequent individuals have also emigrated through family reunification policies.¹⁸ Many continue to emigrate today, although in smaller numbers than in the past. Consequently, emigration still characterizes Turkey for many, despite the fact that the country now receives more immigrants than it sends.¹⁹ 

Turkey’s shift toward immigration can be attributed to a number of factors. Economic growth and higher levels of political stability have led to increased opportunities for labor migrants.²⁰ Its position on the crossroads of two continents has made it a frequent site for transit migrants destined for Europe.²¹ Further, its proximity to areas of conflict in the Middle East has made Turkey a stable and accessible place for those seeking refuge from violence and oppression.²²

Turkey’s recent emergence as a net immigration country does not mean that immigrants are a new phenomenon. Rather, individuals have been settling in Turkey from abroad in significant numbers since the last century of the Ottoman Empire.²³These “old wave” immigrants were often individuals of Turkish origin returning to their ancestral homeland.²⁴ Also included were some non-Turkish Muslim populations who were given non-foreigner status.²⁵ These groups were privileged in the 1934 Law on Settlement as part of a nationalistic push to homogenize the country in terms of ethnicity, language and culture.²⁶ Other populations, including non-Turkish Muslims, Christians, and Jews were subject to forced resettlement and were unable to obtain immigration paperwork.²⁷

However, the reality of global migration began to change toward the end of the Cold War with the emergence of new migration sources. As a result and because of a desire for closer ties to the European Union, Turkey’s immigration rules became less restrictive.²⁸ The current “new wave” of immigration has ushered in the net immigration transition. Most current arrivals are non-Turkish, often non-Muslim, and would have been classified as ineligible “foreigners” under the old Law on Settlement. The initial changes began in the late 1970s when Turkey accepted asylum seekers fleeing the Iranian revolution.²⁹ Flows of refugees continued over the following decades with Iraqis in the 1990s and those fleeing the ongoing Syrian civil war today.³⁰ It is estimated that nearly 2 million Syrians have sought “temporary protection” in Turkey since 2011.³¹ In addition, with economic development, Turkey has received many labor migrants seeking opportunities, particularly from Eastern Europe and the former Soviet Union.³² Labor migrants also arrive from throughout Asia and Africa, but in smaller numbers.³³ Finally, Turkey is host to many retirees from Western Europe, who are attracted by the lower cost of living and the beautiful scenery on the Mediterranean coast.³⁴ 

The diversity in types of immigrants and countries of origin is impressive, with settlers arriving from 176 different countries.³⁵ To put this in perspective, according to data from the UN, the largest single source of immigration to Turkey is Serbia, which sends only 5.88 percent of all immigrants.³⁶ Compare this with the U.S. where Mexico sends 28.3 percent, France where Algeria sends 19.6 percent, or Germany where Turkey supplies 15.7 percent.³⁷ The diverse and global nature of immigrant sources may lead to confusion about the true extent and character of immigration to Turkey.

---

Maintaining a Muslim identity while synthesising western values as a Muslim democracy has resulted in the coining of the term, the “Turkish Model”

---

Estimates vary regarding the actual size of Turkey’s immigrant population.³⁸ The most recent estimate from the World Bank indicates that roughly 1.9 percent of individuals living in Turkey are immigrants.³⁹ According to the U.N., the percentage may be as high as 5.3 percent if one includes refugees and irregular migrants.⁴⁰ Regardless, this is a relatively small number compared to countries that have been receiving immigrants for a long time like the U.S., the U.K., and Germany, where about 13 percent of the population is foreign born.⁴¹ Turkey’s immigrant percentage is more comparable to nations like Japan (2 percent), Hungary (2 percent), Poland (1.75 percent), or perhaps Italy (7 percent) for the U.N. estimate.⁴² Regardless, we currently know little about how knowledgeable ordinary Turks are about their immigrant population, the reasons why they might be wrong, or whether being wrong indicates anything about their attitudes toward immigrants.

Data, Variables, Methodology, and Analytical Sample

Data

I analyze my research questions regarding Turkish innumeracy patterns using data from the 2013 Transatlantic Trends Survey (TATS).⁴³ The data cover individuals aged 18 and over with access to a landline telephone. Respondents were selected via multi-stage probability sampling and have a response rate of 51 percent.⁴⁴ Face-to-face interviewers were conducted between March and July of 2013.⁴⁵ The total sample size was 1002 individuals; however, I focus mainly on the 664 who responded to the innumeracy question.⁴⁶

Variables

The survey measures immigrant population size perceptions with the following question: “In your opinion, what percentage of the total [Turkish] population are immigrants?”⁴⁷ Participants were directed to fill in any number from zero to 100. The responses to this question represent my dependent variable. I describe the Turkish innumeracy patterns in detail in the analysis section.

I also consider the association between innumeracy and various demographic factors. Gender is measured dichotomously. Age is measured with six categories ranging from “18-24” to “65+”. Education is measured with four categories: (1) primary school or less; (2) secondary school degree; (3) college degree; and (4) still studying. Occupation includes five job sector categories: (1) manual laborer; (2) professional; (3) self-employed; (4) service sector employee; and (5) not working. Political party measures the respondent’s voting intention in the next election. I include categories for the two most popular parties in the sample (Adalet ve Kalkınma Partisi and Cumhuriyet Halk Partisi), an “other” category that combines less popular parties,⁴⁸ and a “none” category for those who intend not to vote, to submit a blank ballot, or did not respond. 

To assess where innumeracy is most common, I consider a variable measuring the respondents’ region of the country. The indicator included in the TATS covers seven major geographic regions: (1) Marmara; (2) Aegean; (3) Mediterranean; (4) Central Anatolian; (5) Black Sea; (6) Eastern Anatolian; and (7) South Eastern Anatolian. I also include an eighth category for those living in the major metropolitan areas of Istanbul, Izmir, and Ankara. 

I capture the respondents’ cognitive availability regarding immigrants with a measure of immigrant friendships. The variable has three categories: (1) “no, none”; (2) “yes, a few”; and (3) “yes, many.” Immigrant friendships have proven to be an important predictor of innumeracy in Europe as greater contact with immigrants provides more examples that one can recall when attempting to estimate population size.⁴⁹

I consider 11 measures of attitudes toward immigrants and immigration, which may be related to innumeracy. The first six measure agreement or disagreement with the following statements: (1) immigrants take jobs away from native born Turks; (2) immigrants are a burden on social services; (3) immigrants are a threat to our national culture; (4) immigrants fill jobs with shortages of workers; (5) immigrants create jobs as they set up new businesses; and (6) immigrants enrich our culture. These questions are similar to those used by sociologists trying to measure perceptions of group threat.⁵⁰ Further, I consider two dichotomous variables measuring whether respondents worry about (7) legal immigration and (8) illegal immigration. I also consider (9) the respondents’ perception of the appropriateness of current amount of immigration. The variable has three categories ranging from “too many” to “not too many.” Given Turkey’s unique position as a recent nation of emigration, I consider (10) a dichotomous variable measuring whether respondents agree or disagree that emigration is a problem for Turkey. Finally, I analyze a question assessing (11) the government’s performance in handling immigration. Responses include either “good job” or “poor job.”

Methodology

I begin by describing the patterns of innumeracy in Turkey through a univariate analysis. I then examine how innumeracy varies across my independent variables using bivariate means tables. I assess statistical significance using two sample grouped t-tests. These provide a standard method for gauging whether differences in innumeracy indicate a true pattern in the data or can be attributable to random sampling error. As a standard of significance I use a two-tailed p-value of .05 throughout. All analyses replace the missing values of the independent variables through multiple imputation of 20 datasets.⁵¹ Additionally, all analyses apply sample weights to ensure that the data are representative of the Turkish population.⁵² 

Table 1: Weighted Sample Proportions and Mean Innumeracy Levels – Demographic and Cognitive Availability Variables * indicates a significant (p<.05) difference relative to the reference category.

Analytical Sample

The column titled “Weighted Prop.” In Table 1 provides a basic description of the analytical sample. Of the 664 individuals included, 52 percent are male and 48 percent are female. The sample tends to be younger with 20 percent in the “18-24” category and 25 percent in the “25-34” category. There are less than 10 percent in the “65+” category. Primary school or less is the modal education category with 47 percent of the sample. Given the relatively young sample it is not surprising that nearly 20 percent are still studying. Finally, those who are not working represent the modal occupation category. These include the unemployed, but also retirees, homemakers, students, and the disabled. Of those working the largest percentage (24 percent) are service employees, which includes office clerks, salespeople, nurses, and civil servants. Professionals, which include doctors, lawyers, and managers, are the least common (5 percent).

Analysis

Describing Turkish Innumeracy

The bar chart on the left side of Figure 1 compares the mean estimates of the immigrant population size in Turkey with the actual sizes estimated by the World Bank and the U.N. The typical Turkish respondent believes that the country is 21.2 percent immigrant. This is relative to the actual size estimates of 1.9 and 5.3 percent. Thus, overall Turkish innumeracy is between 15.9 and 19.3 percent. This level of innumeracy is quite extreme, with the higher estimate placing Turkey on par with the U.K. as the third most wildly overestimating nation of the 13 included in the 2013 TATS data.⁵³ Put another way, the typical Turkish respondent perceives the immigrant population between 3.3 and 10 times larger than the reality, depending on the estimate of the truth. 

Figure 1: Perceived Immigrant Population Size in Turkey – Weighted Means and Proportions

To get another perspective on the level of Turkish innumeracy, I divide estimates into categories and present them in the pie chart on the right side of Figure 1. The first category captures those who underestimate, or are roughly correct in their estimate (guesses between 0 and 6.3 percent).⁵⁴ Such individuals constitute a small minority of about 7 percent of the sample. “Overestimators” guess that the immigrant population is between 6.3 and 21.2 percent (the sample mean estimate). These individuals constitute the modal category with 38 percent of the sample. “High Overestimators” guess higher than the sample average and represent 21 percent of the sample. Combined, a clear majority of Turkish respondents, about 60 percent, demonstrates an inflated perception of the immigrant population size. The widespread levels of uncertainty are further evidenced by the one third who did not know and therefore could not provide a response. Innumeracy non-response also represents ignorance regarding the immigrant population size, but in a less interpretable form.⁵⁵

Explaining Turkish Innumeracy

In Table 1, I consider the size estimate means across independent variable categories under the column “Mean Size Estimate.” Consistent with the idea that innumeracy is simply random ignorance, there are few significant differences across the predictors. Size estimates are statistically equivalent across gender, age, occupation, and political party. The unique exception for the demographic variables is education, where those who are still studying offer significantly (t = −3.489; p = .001) lower and more accurate estimates. This may have to do with the recency of study rather than the respondent’s amount of education. Adult students may be more likely to keep track of current social and political events, which could explain their greater accuracy relative to those with college degrees. 

Further, there also appears to be a cognitive availability effect. Although they are rare with only nine percent of the sample, those with many immigrant friends offer significantly (t = −3.569; p = .000) higher immigrant population size estimates. This pattern suggests that regular contact with immigrants makes Turkish respondents perceive immigrants to be more common. Such individuals likely have more available examples of immigrants to use as evidence when formulating estimates. 

Locating Turkish Innumeracy

I consider the geographic distribution of size misperceptions in Figure 2, which maps mean size estimates in the eight Turkish regions considered in the TATS. Areas are color-coded such that lower mean estimates are lighter and higher estimates darker. The metropolitan region (Istanbul, Izmir, and Ankara) acts as the reference category. This region, along with Marmara and the Black Sea, offer the lowest and most accurate size estimates. The former is particularly interesting given that Istanbul has the largest concentration of immigrants in the country.⁵⁶ Based on cognitive availability, one might assume that those living in urban areas where immigrants tend to settle would offer the highest estimates.⁵⁷ Rather, those in the Central Anatolian and Mediterranean regions offer the highest and least accurate estimates. The latter is the most popular among European retirees, particularly the city of Antalya.⁵⁸ Further, it, along with the Southeastern Anatolian region, have the highest concentrations of Syrian refugee camps.⁵⁹ These individuals may be more visible to ordinary Turks, a factor which produces larger size estimates. Both the Mediterranean and Central Anatolian regions, as well as the moderately innumerate Aegean and Southeast Anatolian regions, offer mean estimates that are significantly (p<.05) higher than the metropolitan region.

Figure 2: Innumeracy across Turkish Regions

Potential Consequences of Turkish Innumeracy

The connection between immigrant population size estimates and attitudes toward immigrants have been particularly intriguing to social researchers. Generally, the pattern in the U.S. and Europe is that the more inflated one’s perception of the immigrant population, the more negatively he or she will perceive immigrants. I consider this possibility with my Turkish sample in Table 2. 

Table 2: Weighted Proportions and Mean Innumeracy Levels – Attitudes toward Immigrants/Immigration

 * indicates a significant (p<.05) difference relative to the reference category.

First, the column titled “Weighted Prop.” provides a basic description of Turkish attitudes toward their immigrant population. Overall, despite their being a small population, the typical Turkish respondent does not view immigrants positively. Nearly three-quarters agree that immigrants take jobs from natives and are a burden on social services. Further, nearly 60 percent view immigrants as a threat to Turkish culture. Clear majorities disagree that immigrants create jobs and enrich Turkish culture. However, a small majority concedes that immigrants fill jobs with shortages of workers. Most Turks are worried about both legal and illegal immigration, and they think their government has done a poor job of handling these populations. Despite a general negativity, only about 27 percent of Turkish respondents feel that there are too many immigrants in the country. This percentage is the sixth lowest relative to other countries included in the TATS.⁶⁰ Finally, about half believe that emigration is a problem.

---

Overestimating the immigrant population does not make respondents more threatened by immigrants or more worried about immigration in general

---

Interestingly, there appears to be no evidence for the consistent innumeracy-attitudes association that has been observed in Europe and the U.S. Conversely, Turks who view immigrants more positively as creating jobs (t = −2.106; p = .036) and filling jobs with work shortages (t = −4.118; p = .000), actually offer significantly more inflated size estimates compared to those with more negative perceptions. Further, those worried about legal immigration express significantly (t = −3.167; p = .002) lower and more accurate size estimates than those expressing no concern. These patterns run counter to commonly held assumptions about innumeracy and suggest that Turkey represents a unique case relative to other contexts. Finally, those who offered lower and more accurate size estimates are also significantly more likely to think that emigration is not a problem (t = 5.959; p = .000) and that the government is doing a poor job handling immigration (t = 2.192; p = .029).

Discussion and Conclusion

As Turkey emerges as a destination for immigrants more so than a source of emigrants, it becomes important to understand how the country is adapting to its growing immigrant population. The current study examined how accurately Turkish citizens perceive their immigrant population. Research on such perceptions throughout the U.S. and Europe have revealed high levels of ignorance as well as potential consequences if inflated perceptions become widespread. The preceding analysis offers multiple findings and implications for future research, which I highlight below. 

Innumeracy is Quite High in the Turkish Context

Overall, there are high levels of immigrant population innumeracy among Turkish respondents. This follows previous research demonstrating high levels throughout Europe and the U.S. It is also somewhat counter to Düvell’s assumption that ordinary Turkish citizens are not aware of the transition to net immigration.⁶¹ While the current data do not directly prove widespread awareness of the actual transition, they do suggest that ordinary Turks view their immigrant populations as being much more substantial than the reality.

To put these numbers in context, one can compare them to the results from countries with comparable immigrant population sizes. In the 2013 TATS, the closest matches for Turkey are Poland (about 2 percent immigrants) and Slovakia (about 4 percent immigrants).⁶² On average, respondents overestimate by ten percent in the former and five percent in the latter. Turkey’s 19.3 percentage points of incorrectness on average places the country on par with the U.K. (19 percentage points incorrect on average), which has a much larger immigrant population and a more established history as a nation of immigration.⁶³

Innumeracy in Turkey is not Completely Random

If independent variables yield no statistically significant associations with innumeracy, it would suggest that these misperceptions are simply random ignorance. While the current study found few such associations, the existence of some indicates that Turkish innumeracy is not completely random. The lowest levels of innumeracy were found among those still in school. This is likely due to current students being more recently exposed to politics and national demographics through their studies. The highest levels of innumeracy were found among those living in Central Anatolia and the Mediterranean region and among those claiming to have many immigrant friends. The latter follows hypotheses from the previous innumeracy literature regarding cognitive availability.⁶⁴ Those with immigrant friends have more personal examples of immigrants that come to mind when attempting to formulate an estimate.

---

The long history of emigration has made Turks more understanding of those seeking opportunities abroad and more willing to accept an immigrant presence despite negative attitudes

---

Unfortunately, the current study can only scratch the surface of possible factors that might account for Turkish innumeracy. Being limited to the questions included in the 2013 TATS, I am unable to consider measures of immigrant neighbors, immigrant co-workers, or immigrant classmates, which may all be additional sources of cognitive availability. Further, previous innumeracy research highlights media exposure as another important source of information. Where do respondents get their information about immigrants: newspapers, television, the internet, or some other source? Previous research suggests that the former often produces the most accurate perceptions.⁶⁵ Overall, more data and research are needed to understand Turkish innumeracy patterns completely.

Innumeracy in Turkey Appears to be Innocuous 

Despite ranking among the most innumerate countries, innumeracy among Turkish respondents was unrelated to negative attitudes toward immigrants. Overestimating the immigrant population does not make respondents more threatened by immigrants or more worried about immigration in general. The opposite was true in some cases. This sets Turkey apart from other nations as this pattern is contrary to most of the existing research on innumeracy from the U.S. and Europe.⁶⁶ I do note that this unexpected pattern should be interpreted with caution until it can be confirmed with additional data from the Turkish context.

Regardless, the lack of an association between innumeracy and negative attitudes does not mean that Turkish respondents are completely comfortable with immigrants. On the contrary, the current data indicate that feelings of threat from immigrants and concern about immigration are common. However, Turks display a degree of ambivalence toward foreigners as, despite the negativity and the inflated perceptions, relatively few are willing to declare that there are too many immigrants. Perhaps the long history of emigration has made Turks more understanding of those seeking opportunities abroad and more willing to accept an immigrant presence despite negative attitudes.

It is also possible that more time is needed for the innumeracy and anti-immigrant sentiment relationship to fully crystallize. Turkey’s immigrant population remains small and relatively recent. The paradigm shift discussed by Düvell is still in its early stages. Taken together, patterns may change as foreign-born populations grow and Turkey begins to view itself as an immigrant destination. Intergroup contact between immigrants and natives should increase with time. Intergroup competition for jobs and neighborhoods, which can engender negative attitudes, may increase as well. Both of these are related to innumeracy in other contexts and may need time to develop. Thus, it is necessary for researchers to continue monitoring innumeracy and its connection to attitudes in Turkey as it emerges on the world stage as an immigrant destination. The current study is an important first step, but continuing with such an endeavor may elucidate why innumeracy is so often linked to dislike of immigrants.

High levels of innumeracy and of dislike for immigrants are troubling even if they are not currently linked. It suggests a general level of ignorance and discomfort regarding immigrants in Turkey that may worsen in the future. If anti-immigrant political parties and claims-makers begin to emerge, as has occurred in Europe, a highly innumerate population may be easily convinced by their arguments and willing to use immigrants as scapegoats for social problems. It is important for policy makers to be aware of this possibility, but it will require further research and monitoring from social scientists.

A final possibility and another avenue for future research lies in the diversity of Turkey’s immigration sources. With so many sources one must question whom Turks have in mind when they estimate. European retirees? Laborers from Eastern Europe? Refugees from Syria? Is any particular group more or less threatening than the others? For example, it is possible that if respondents picture refugees from neighboring Syria, they may be more sympathetic. Undoubtedly, such a large influx of foreigners over a short period of time will affect how Turks view immigrants, likely generating negative attitudes⁶⁷ and misperceptions about size of the foreign-born population. However, Syrian refugees are fleeing circumstances outside of their control, are intended to be temporary in status, and share a religion with the host society. This could explain why innumeracy is high, but overestimation is unrelated to negative attitudes. However, the vast array of possible groups that respondents could have in mind makes innumeracy difficult to interpret on its own. We may learn more if surveys begin to ask ordinary Turks where they think most immigrants are coming from and why they are here, along with their size perceptions. Ignorance regarding the qualitative character of immigration has proven informative in the U.K.⁶⁸ and Finland,⁶⁹ and may also help us to understand the Turkish context.

Conclusion

Turkey’s transition to net immigration is recent and their foreign born population is small. Despite this, if you ask the typical Turkish individual, they will tell you that immigrants represent one fifth of their country. The seemingly ubiquitous phenomenon of immigrant population innumeracy is confirmed to exist in Turkey, but seems to be relatively innocuous. Whether this will remain true as Turkey asserts an identity as an immigrant receiver will be an interesting task for future research to determine. At some point innumeracy may begin to link to negative attitudes in Turkey, as it has in many other contexts. Determining if and how this occurs will help researchers to better understand the phenomenon of innumeracy in general. 

Endnotes

1. The World Bank, “Net Migration,” retrieved July 16, 2015, from http://data.worldbank.org/indicator/SM.POP. NETM; Franck Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift,” 
   Insight Turkey, Vol. 16, No. 4 (2014), pp. 87-108; Juliette Tolay, “Discovering Immigration into Turkey: The Emergence of a Dynamic Field,” International Migration, IOM doi:10.1111/j.1468-2435.2012.00741.x (2012), p. 8.
2. Russell King, Anthony Fielding, and Richard Black, “The International Migration Turnaround in Southern Europe.” R. King, A. Fielding and R. Black (eds), Southern Europe and the New Immigrants, (Portland, 1997).
3. Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift,” p. 83.
4. Ibid, p. 87.
5. Jack Citrin and John Sides, “Immigration and the Imagined Community in Europe and the United States,” Political Studies, Vol. 56, No. 1 (2008), pp. 33-56; Daniel Herda, “‘How Many Immigrants?’ Foreign Born Population Innumeracy in Europe,” Public Opinion Quarterly, Vol. 74, No. 4 (2010), pp. 674-695; Daniel Herda, “Too Many Immigrants?: Examining Alternative Forms of Immigrant Population Innumeracy,” Sociological Perspectives, Vol. 56, No. 2 (2013), pp. 213-240; Mikael Hjerm, “Do Numbers Really Count? Group Threat Theory Revisited,” Journal of Ethnic and Migration Studies, Vol. 33, No. 8 (2007), pp. 1253-75; Moshe Semyonov, Rebeca Raijman, and Anastasia Gorodzeisky, “Foreigners’ Impact on European Societies: Public Views and Perceptions in a Cross-National Comparative Perspective,” International Journal of Comparative Sociology, Vol. 49, No. 3 (2008), pp. 5-29; Moshe Semyonov, Rebeca Raijman R, Anat Yom Tov and Peter Schmidt, “Population Size, Perceived Threat and Exclusion: A Multiple-Indicators Analysis of Attitudes Toward Foreigners in Germany,” Social Science Research, Vol. 33, No. 4, (2004), pp. 681-701; John Sides and Jack Citrin, “European Opinion about Immigration: The Role of Identities, Interests and Information,” British Journal of Political Science, Vol. 37, No. 3 (2007), pp. 477-504; Charles A. Gallagher, “‘Blacks, Jews, Gays and Immigrants are Taking Over’: How the Use of Polling Data Can Distort Reality and Perpetuate Inequality among Immigrants,” Ethnic and Racial Studies, Vol. 37, No. 5 (2014), pp. 731-737.
6. Ipsos MORI, “Perceptions Are Not Reality: Things the World Gets Wrong,” retrieved July 16, 2015, from https://www.ipsos-mori.com/researchpublications/researcharchive/3466/Perceptions-are-not-reality-Things-the-world-gets-wrong.aspx
7. Ibid.
8. Ibid.
9. In addition to misperceptions about population size, other research has documented uncertainty about various qualitative characteristics of immigrant populations. Herda demonstrated that nearly one-fifth of Finnish respondents faultily perceived Somalia to be the primary source of immigrants in Finland, rather than the correct answer: Russia. Further, Blinder found that respondents in the U.K. were more likely to believe that immigrants were labor migrants or asylum seekers, rather than the correct answer: students. These qualitative misperceptions, while interesting, are beyond the scope of the current analysis. Daniel Herda, “Beyond Innumeracy: Heuristic Decision-Making and Qualitative Misperceptions about Immigrants in Finland,” Ethnic and Racial Studies, Vol. 38, No. 9 (2015), pp. 1627-1645; Scott Blinder, “Imagined Immigration: The Impact of Different Meanings if ‘Immigrant’ in Public Opinion and Policy Debates in Britain,” Political Studies, Vol. 63, No. 1 (2013), pp. 80-100.
10. Ibid.
11. Richard Nadeau, Richard G. Niemi, and Jeffrey Levine, “Innumeracy about Minority Populations,” The Public Opinion Quarterly, Vol. 57, No. 3 (1991), pp. 332-47.
12. Nadeau, Niemi, and Levine, “Innumeracy about Minority Populations”; Richard Alba, Ruben G. Rumbaut, and Karen Marotz, “A Distorted Nation: Perceptions of Racial/Ethnic Group Sizes and Attitudes toward Immigrants and Other Minorities,” Social Forces, Vol. 84, No. 2 (2005), pp. 901-19; Herda, “‘How Many Immigrants?’ Foreign Born Population Innumeracy in Europe.”
13. Herda, “Too Many Immigrants?: Examining Alternative Forms of Immigrant Population Innumeracy.”
14. Amos Tversky and Daniel Kahneman, “Availability: A Heuristic for Judging Frequency and Probability,” Cognitive Psychology, Vol. 5, No. 2 (1973), pp. 207–232.
15. From Daniel Herda, “‘How Many Immigrants?’ Foreign Born Population Innumeracy in Europe” and Daniel Herda, “Too Many Immigrants?: Examining Alternative Forms of Immigrant Population Innumeracy.”
16. Citrin and Sides, “Immigration and the Imagined Community in Europe and the United States”; 
    Herda, “Too Many Immigrants?: Examining Alternative Forms of Immigrant Population Innumeracy”; Hjerm, “Do Numbers Really Count? Group Threat Theory Revisited”; Semyonov, Raijman, and Gorodzeisky, “Foreigners’ Impact on European Societies: Public Views and Perceptions in a Cross-National Comparative Perspective”; Semyonov, Raijman R, Yom Tov and Schmidt, “Population Size, Perceived Threat and Exclusion: A Multiple-Indicators Analysis of Attitudes toward Foreigners in Germany.”
17. Susan Beth Rottmann, “Negotiating Modernity and Europeanness in the Germany-Turkey Transnational Social Field,” Insight Turkey, Vol. 16, No. 4 (2014), pp. 148-158.
18. Ibid.
19. The World Bank, “Net Migration.”
20. Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift.”
21. Ibid.
22. Ibid.
23. Tolay, “Discovering Immigration into Turkey: The Emergence of a Dynamic Field.”
24. Ibid.
25. Erol Ülker, “Assimilation, Security and Geographical Nationalization in Interwar Turkey: The Settlement Law of 1934,” European Journal of Turkish Studies, Vol. 7 (2008), http://ejts.revues.org/21123
26. Ibid.
27. Ibid.
28. Juliette Tolay, 2012, “Turkey’s “Critical Europeanization”: Evidence from Turkey’s Immigration Policies,” pp. 39-62, Seçil Paçacı Elitok and Thomas Straubhaar (eds.), Turkey, Migration and the EU: Potentials, Challenges and Opportunities, Edition HWWI, No. 5, ISBN 978-3-937816-94-4, http://nbn-resolving.de/urn:nbn:de:gbv:18-3-1187
29. Tolay, “Discovering Immigration into Turkey: The Emergence of a Dynamic Field.”
30. M. Murat Erdoğan, “Perceptions of Syrians in Turkey,” Insight Turkey, Vol. 16, No. 4 (2014), pp. 65-75; Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift.”
31. Ibid.
32. E. M. Özgür, A. Deniz, D. Hasta, M. M. Yüceşahin and S. Yavuz, “The Immigration of Russians and Azerbaijanis to Turkey: Who are They? Why are They Here? Insight Turkey, Vol. 16, No. 4 (2014), pp. 105-121.
33. Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift.”
34. Canan Balkır and İlkay Südaş, “Guests and Hosts: European Retirees in Coastal Turkey,” Insight Turkey, Vol. 16, No. 4 (2014), pp. 123-142; Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift.”
35. Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift.”
36. Justin Gest and Anna Boucher, “Crossroads of Migration Stock Map,” Harvard University Center for Geographic Analysis, retrieved July 16, 2015, from www.crossroadsofmigration.com. 
37. Ibid.
38. I note that irregular migration patterns taken by many immigrants to Turkey and inconsistent record keeping by Turkish officials makes it difficult to accurately estimate the actual foreign born population. Thus, I include two different estimates of the truth. Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift.”
39. 39. World Bank, Turkey, Migration and Remittances Factbook (Washington: World Bank, 2010), retrieved July 12, 2014, from http:// siteresources.worldbank.org/INTPROSPECTS/Resources/334934-1199807908806/Turkey.pdf
40. United Nations, Migrant stock (Washington: UN Population Division, 2013), retrieved July 16, 2015, from http://esa.un.org/ unmigration/TIMSA2013/Data/subsheets/UN_MigrantStock_2013T3.xls
41. Ipsos MORI, “Perceptions Are Not Reality: Things the World Gets Wrong.”
42. Ibid.
43. German Marshall Fund of the United States (GMF), Washington, USA: Isernia, Pierangelo (2014): Transatlantic Trends 2013. GESIS Data Archive, Cologne. ZA5646 Data file Version 1.0.0, doi: 10.4232/1.11894
44. Pierangelo Isernia, “Codebook Transatlantic Trends Survey 2013,” ZA6546: Transatlantic Trends, (2013).
45. Ibid. Note that data collection was suspended briefly during this period amid nationwide protests.
46. Compared to the analytical sample, non-responders are more likely to be female (42.7 percent v. 58.7 percent) and to disagree that immigrants create jobs (56.0 percent v. 43.7 percent), and are less likely to graduate secondary school (15.2 percent v. 25.5 percent) and be manual laborers (12.32 percent v 19.7 percent). All other independent variable comparisons are statistically equivalent across responders and non-responders.
47. Isernia, “Codebook Transatlantic Trends Survey 2013.”
48. The largest categories are the Milliyetçi Hareket Partisi (9.68 percent) and the Barış ve Demokrasi Partisi (3.69 percent). All other parties had only six or fewer respondents indicating support. If I do not collapse these categories, the results, as those presented below, show no significant differences in innumeracy levels across political party.
49. Herda, “Too Many Immigrants?: Examining Alternative Forms of Immigrant Population Innumeracy.”
50. Under group threat theory prejudice develops when a dominant group perceives its social position and control over resources (e.g.: jobs, education, and housing) to be threatened by minority encroachment. See Herbert Blumer, “Race Prejudice as a Sense of Group Position,” Pacific Sociological Review, Vol. 1, No. 1 (1958), pp. 3-7. Lawrence Bobo, “Whites’ Opposition to Busing: Symbolic Racism or Realistic Group Conflict?” Journal of Personality and Social Psychology, Vol. 45, No. 6 (1983), pp. 1196-210.
51. Over 83 percent of cases have no missing values. The variables with the highest percentages with missing observations are the measures of the government’s handling of immigration (6.7 percent) and emigration is a problem (6.1 percent).
52. Isernia, “Codebook Transatlantic Trends Survey 2013”.
53. The German Marshall Fund, “Transatlantic Trends: Key Findings 2013,” accessed July 21, 2015, from http:// trends.gmfus.org/files/2013/09/TTrends-2013-Key-Findings-Report.pdf
54. I set the maximum estimate to be considered correct at a single percentage point above the UN estimate. This follows Herda’s operationalization. Herda, “Too Many Immigrants?: Examining Alternative Forms of Immigrant Population Innumeracy.”
55. Ibid.
56. Balkır and Südaş, “Guests and Hosts: European Retirees in Coastal Turkey.”
57. While logical, there does not seem to be much evidence for an urbanicity effect in the existing literature. See Daniel Herda, “‘How Many Immigrants?’ Foreign Born Population Innumeracy in Europe.”
58. Balkır and Südaş, “Guests and Hosts: European Retirees in Coastal Turkey.”
59. UNHCR, “Syrian Refugees in the Region,” accessed August 26, 2015 http://reliefweb.int/sites/reliefweb.int/files/ resources/SyrianREfugees intheRegion2014Jan07-.pdf
60. The German Marshall Fund, “Transatlantic Trends: Key Findings 2013”.
61. From Franck Düvell, “Turkey’s Transition to an Immigration Country: A Paradigm Shift.”
62. The German Marshall Fund, “Transatlantic Trends: Key Findings 2013.”
63. Ibid.
64. Herda, “Too Many Immigrants?: Examining Alternative Forms of Immigrant Population Innumeracy.”
65. Herda, “‘How Many Immigrants?’ Foreign Born Population Innumeracy in Europe.”
66. While such a finding is rare, it is not without precedent. In Herda’s study of innumeracy regarding the Jewish population in the U.S., high levels of innumeracy were unrelated to anti-Semitic attitudes. Daniel Herda, “Innocuous Ignorance: Perceptions of the American Jewish Population Size,” Contemporary Jewry, Vol. 33, No. 3 (2013), pp. 241-255.
67. Deniz Cengiz, “Zorunlu Göçün Mekânsal Etkileri Ve Yerel Halkın Algısı Kilis Örneği,” Turkish Studies, Vol. 10, No. 2 (2015), pp. 101-122.
68. Blinder, “Imagined Immigration: The Impact of Different Meanings of ‘Immigrants’ in Public Opinion and Policy Debates in Britain.”
69. Herda, “Beyond Innumeracy: Heuristic Decision-Making and Qualitative Misperceptions about Immigrants in Finland”