Thanks, But No Thanks: Attitudes On Refugee Policy in the European Union

Interdisciplinary Seminar in Empirical Social Science

21 June 2022

Natalia Letki & Dawid Walentek

University of Warsaw

About us

• Natalia Letki Principal Investigator (University of Warsaw)
• Dawid Walentek (University of Warsaw)
• Artem Graban (University of Warsaw)
• Ulf Liebe (University of Warwick)
• Peter Thisted Dinesen (University College London)
   
• Funding: NCN (National Science Centre Poland) grant no 2019/33/B/HS6/00841
   
• Our website is here

Background

• 2015 Refugee Crisis
• Budapest's, Prague's and Warsaw's refusal to join the relocation scheme
• Refugee policy in the EU
  is a patchwork
• Ongoing relevance: crisis on the Polish-Belarussian border, war in Ukraine

Literature

• Preferred refugee profiles and sentiments towards
  out-group members  (e.g. Adida, Lo, and Platas 2019)
• Effects of exposure on acceptance of out-group members (e.g. Schaub, Gereke, and Baldassarri 2020)
• Behavioural and attitudinal changes in relation to out-group members (e.g. Alesina, Miano, and Stancheva 2018)
• Limited research on refugee policy:
  • ​Only interested in allocation (Bansak, Hainmueller, and Hangartner 2017)
  • Asks about policy solutions the EU cannot deliver  (Jeannet, Heidland, Ruhs 2021)
  • ​Looks at general statements, e.g. 'government should be generous in judging  asylum applications' (Van Hootegem, Meuleman and Abts 2020).

Our project

• We set out to identify what refugee policy Europeans want
• We study attitudes in respect to:
  • allocational regime
  • level of border control
  • right to work
  • freedom of movement
  • cost of the policy
• We test the correspondence between attitudes and behaviour
• We look at the effect of informational and visual priming
  on attitudes and behavior

Data & Methods

Data

• Online survey experiment
  in early 2022
• 10 EU Member States
• 16,976 respondents
• 4 treatment groups
• 3 components
  • conjoint
  • real-effort
  • questionnaire
• Follow up study (DE, PL & HU) in spring 2022
• Preregistration with EGAP

Treatments

• Three visual primes
  • Welfare
  • Security
  • Humanitarian
• Images accompanied by a short caption
• Widely circulated in media in relation to the relevant primes
• Text-based Status Quo prime on current refugee policy in a given country

Methods: Conjoint

• Full randomisation
• Five attributes over twelve levels resulting in 72 policy profiles
• Respondents make six choices over a pair of profiles
• Over 101,856 (force)choices made and 203,712 profile ratings assigned
• Each profile appears on average 2829 times

Methods: Real effort

• Final part of the survey
• Respondents receive first general information about the International Rescue Committee
• Then respondents are presented with the real-effort task, where they must decide whether to donate half of the survey remuneration to the IRC

Results

Results: AMCE vs MM

• Average Marginal Component Effects (AMCE) - the impact of attribute level on favourability of a refuge policy profile, relative to a given attribute's baseline (Hainmueller et al. 2014)
• Marginal Means (MM) - allow to produce descriptive inference about differences in preferences across subgroups in a multidimensional experimental setting (Leeper et al. 2020)

Conjoint: control group

• No meaningful role of the allocation regime
• Preference for increased border control
• Preference for limited freedom of movement
• Preference for the right to work (and dislike of the alternative)
• Dislike of the costly policy and preference for the low costs

Country heterogeneity

Country heterogeneity

• Strong variation between countries in respect to the allocation regimes
  • DE, ES, AU & DEN dislike the Status Quo while HU & SL prefer it
  • DE, ES, AU & PT support relocation while HU opposes
  • All indifferent on fiscal solidarity bar supportive DEN & negative PT
  • Note: DE, ES & AU receive the most and HU the least asylum applications
• Hardly any meaningful variation between the Member States in respect to other attributes than allocation regimes

Ideological heterogeneity

• Ideology: 0 to 10 scale, Left to Right, recoded
• Right more restrictive,
  Left more favorable
• BUT Left against freedom of movement and fiscal solidarity

Real effort

• Control group:
  • PL, DE & BG ~ 25%
  • HU & SL ~ 30%
  • AU & HR ~ 35%
• Treatments:
  • Humanitarian only
  • ATE = .05
    5% higher donation
  • Bonferroni
    correction
  • Region FE (NUTS 2)

Real effort: heterogeneity

• Meaningful heterogeneity in the treatment effect
  for different levels
  of ethnocentrism
• Treatment effect
  does not set-in any more for higher levels
  of ethnocentrism
• Ethnocentrism
  based on a set of survey questions

Real effort: attitudes
and behaviour

• We see a correspondence between attitudes
  and behaviours
• Average rating assigned to profiles
  with a relocation
  attribute level

Panel

• Three countries in April & May
  • Germany, Poland and Hungary
• Two samples
  • recontact with January & February round (problematic)
  • fresh sample
• Motivated by the onset of the war in Ukraine

Panel: policy preferences 

• Consistent preferences
  & ratings of profiles
• Stronger sentiments about refugees joining the labour market
• Change in attitudes towards freedom
  of movement

Panel: Real Effort Task

• Lower level
  of donation
• Greater effect size
  in the recontact group

Discussion

What do we learn?

• Disagreement on allocation: Smart Solidarity likely to turn out
  to be a disappointment, while dissatisfaction in DE, ES & AU may grow
• Agreement on right to work, freedom of movement, border control and policy cost:
  • Welcoming on the labour market, but not on the streets
  • Concerned about costs and borders
     
• Responsive to humanitarian media frame, but not across the (ethnocentric) board
• Attitudes and behaviours do seem to overlap
   
• Preferences stable, despite the war in Ukraine
• Status Quo bias on movement & right to work
• Crowding out effect on donations  

Thank you