Do attitudes expressed in surveys predict ethnic discrimination?

ABSTRACT

Survey data on people’s reported attitudes towards ethnic minorities are sometimes used as a proxy for ethnic discrimination. However, there is weak empirical evidence of a link between reported attitudes and discrimination. In this article, we use survey data on people’s attitudes towards ethnic minorities combined with a direct measure of ethnic discrimination from a field experiment in the Swedish housing market to re-examine this policy-relevant issue. We find clear evidence of a link between reported attitudes towards ethnic minorities and the extent of ethnic discrimination: in regions where attitudes are more negative, there is more discrimination, and vice versa. Thus, in contrast to most prior studies, our results suggest that reported attitudes may be a useful predictor of ethnic discrimination.

KEYWORDS:

• Attitudes
• ethnic discrimination
• housing market
• field experiments
• surveys
• racial prejudice

Acknowledgements

We are grateful for helpful comments from seminar participants at the ESPE annual congress in Berne, Stockholm University, and Uppsala University, and colleagues at the Linnaeus University Centre for Labor Market and Discrimination Studies.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCiD

Magnus Carlsson http://orcid.org/0000-0002-5620-4745

Notes

1 Another common approach to studying ethnic discrimination is to use administrative data and measure discrimination as the remaining differences across ethnic groups after taking into account differences in all observable factors (Altonji and Blank 1999). However, this approach has been severely criticized, and currently, most discrimination studies use the correspondence testing approach, that is, sending fictitious written applications to employers (cf. Riach and Rich 2002; Dymski 2006; Rich 2014). A drawback with these experimental studies is that they are costly and time-consuming to perform, and hence, other types of measures – such as surveys – are often needed to guide policy formulation.

2 Two other examples are Kutner, Wilkins, and Yarrow (1952) and Saenger and Gilbert (1950). There are also a few non-experimental studies that examine the relationship between attitudes and labour market outcomes combining survey and administrative data (e.g. Åslund and Rooth 2005; Charles and Guryan 2008; Waisman and Larsen 2008; Carlsson and Rooth 2016). The results of these studies vary, but a problem with this approach is that it is difficult to determine if an estimated ethnic difference in labour market outcomes reflects ethnic discrimination or ethnic differences in unobserved characteristics.

3 The experiment was conducted for a larger research project on age, gender, and ethnic discrimination in the Swedish housing market (cf. Carlsson and Eriksson 2014). The design of the experiment uses a similar methodology as Bertrand and Mullainathan (2004) and Eriksson and Rooth (2014).

4 This topic is discussed in, for example, Eriksson (2010).

5 The other top-ten countries are Hungary (4,188), Austria (3,323), Denmark (2,656), Switzerland (2,750), Norway (2,474), Germany (2,115), Luxembourg (1,806), the Netherlands (1,421), and Belgium (1,238); cf. Table 1.5 in OECD (2015).

6 This is the case despite the fact that Sweden, in an international comparison, offers newly arrived refugees and their relatives a very extensive integration program with language courses, other training, internships, etc.

7 Cf. European Social Survey (2015).

8 This topic is discussed in, for example, Rydgren (2007) and Rydgren and Ruth (2013).

9 See Rich (2014) for a survey and Ahmed and Hammarstedt (2008) for an example of a study.

10 The website is called Blocket (www.blocket.se).

11 Typically, these landlords have their own queuing systems.

12 For men, the typical Swedish name was Erik Johansson and the typical Arabic/Muslim name was Ali Hassan. For women, the corresponding names were Anna Nilsson and Fatima Ahmed. For each group, these are among the most common names in Sweden, according to Statistics Sweden’s name register. This experiment also randomly varied other attributes than the name. How these other attributes affect the landlords’ responses do not affect our estimates of ethnic discrimination because all attributes are randomly assigned. Except for the name of the applicant, five other attributes were randomly assigned to each inquiry: age, employment status, leisure time interests, whether the applicant was a smoker, and whether the applicant had a reference from a previous landlord. The age of the applicant was between 25 and 55; employment status was either unemployed, employed as a shop sales assistant, or employed as a financial manager; leisure time interests were either enjoying evenings at home, being engaged in sport activities, or enjoying restaurant life and nightclubs; either the applicant was a smoker (who never smoked indoors) or a non-smoker; references was either a sentence that expressed that the applicant had a reference from a previous landlord or no information about references.

13 Landlords who explicitly stated that they preferred to be contacted only by phone were discarded.

14 FSI is a Swedish research institute that measures various attitudes of the Swedish population. The FSI attitude survey has been conducted yearly since the 1980s. The response rate is approximately sixty per cent. SOM at the University of Gothenburg has conducted a yearly survey since 1986. The response rate is approximately seventy per cent.

15 We do not have access to the 2011 survey.

16 The mean is greater than the median because there are a few municipalities with a large number of survey respondents. There are many respondents in these municipalities because they have a large population and both surveys use representative sampling.

17 We have compared the characteristics of the respondents in the FSI and SOM surveys and the landlords in the field experiment in the dimensions where this is possible, that is, where comparable characteristics are available. In the field experiment, fifty-five per cent of the landlords have a male name and thirteen per cent have a foreign-sounding name. In the FSI (SOM) survey data, forty-seven (forty-eight) per cent of the respondents are men and seventeen (five) per cent are born outside Sweden (note that some individuals with foreign-sounding names may be born in Sweden).

18 We exclude the municipalities in Stockholm County, Gothenburg municipality, and Malmo municipality.

19 The FSI survey also contained a few other questions about immigration, but these questions focus more on immigration legislation. The SOM survey did not include any other relevant questions.

20 The five possible answers are “Very valuable”, “Quite valuable”, “Not very valuable”, “Not valuable at all”, and “Unsure, do not know”.

Additional information

Funding

This work was supported by the Jan Wallander and Tom Hedelius Foundation [grant number W2010-0049:1].