Understanding inequalities in COVID-19 vaccination between migrants and non-migrants in Germany: The role of psychological factors of vaccine behaviour

This study analyzes German data to reveal that psychological factors, particularly vaccine confidence, explain nearly two-thirds of the vaccination gap between migrants and non-migrants, though the specific drivers vary significantly by country of origin, necessitating tailored health communication strategies.

The Gist

Imagine Germany as a massive, bustling city trying to protect its residents from a invisible storm (the virus). To do this, the city offers a special "umbrella" (the vaccine) to everyone for free. Most people grab an umbrella quickly, but the study found that some groups of people—specifically those who moved to Germany from other countries (migrants)—were holding onto their umbrellas a bit longer or not picking them up at all.

This paper is like a detective story trying to figure out why this gap exists. The researchers didn't just look at who didn't have an umbrella; they looked at what was going on in their heads that made them hesitate.

Here is the breakdown of their investigation using simple analogies:

1. The Mystery: The "Umbrella Gap"

The researchers found a clear gap. People born in Germany (non-migrants) had a very high vaccination rate (about 94%). However, people who moved there themselves (first-generation migrants) and their children born in Germany (second-generation) had lower rates (around 88-89%).

The gap was even wider for people coming from specific regions:

• Eastern Europe: A noticeable gap.
• Middle East & North Africa (MENA): The biggest gap. Only about 80% were vaccinated.

2. The Tool: The "5C Backpack"

To understand the hesitation, the researchers used a mental checklist called the 5C Model. Imagine every person carries a backpack with five specific items that decide if they will get vaccinated:

1. Confidence: Do I trust the umbrella? (Is it safe? Do I trust the people making it?)
2. Complacency: Do I even need an umbrella? (Is the storm really dangerous?)
3. Constraints: Is it too hard to get the umbrella? (Is it too far away? Too confusing?)
4. Calculation: Am I doing the math? (Is the risk of the storm worth the risk of the umbrella?)
5. Collective Responsibility: Do I want to protect my neighbors? (Is this a community effort?)

3. The Big Discovery: It's Mostly About "Trust"

The researchers ran the numbers to see how much of the "Umbrella Gap" could be explained by these five mental items.

The Result: About two-thirds (61-64%) of the reason why migrants were less vaccinated than non-migrants was due to these psychological factors.

The most powerful factor? Confidence.

• Migrants, on average, had less trust in the safety of the vaccine and less trust in the health system compared to non-migrants.
• They also felt less "Collective Responsibility" (a sense of belonging to the community effort).
• Interestingly, some migrants felt the storm wasn't a big deal (Complacency) or spent too much time worrying about the pros and cons (Calculation), which made them hesitate.

4. The Twist: Not All Migrants Are the Same

This is where the story gets really interesting. The "5C Backpack" explained the hesitation differently depending on where the person came from.

• For people from Eastern Europe: The backpack explained 73.5% of the gap. It turns out, their hesitation was mostly about what was in their heads (distrust, thinking it wasn't necessary, etc.). If you fixed their trust and information, you would likely fix the vaccination gap.
• For people from the Middle East & North Africa (MENA): The backpack only explained 19.4% of the gap. This is a huge clue! It means that for this group, the "mental" reasons (distrust, etc.) are only a small part of the problem. The rest of the problem is likely structural barriers—things like language difficulties, fear of legal trouble, not knowing where to go, or feeling discriminated against. The "umbrella" might be there, but the path to get it is blocked by walls that the 5C model can't see.

5. The Conclusion: How to Fix the Gap

The study concludes that we can't just use a "one-size-fits-all" approach to get everyone vaccinated.

• For the "Trust" groups (like Eastern Europe): We need better communication. We need to build trust, explain the science clearly, and show that the vaccine is safe. We need to fill their "Confidence" backpack.
• For the "Barrier" groups (like MENA): Talking isn't enough. We need to remove the walls. We need multilingual guides, community leaders they trust, and clear assurances that getting vaccinated won't get them in trouble with immigration authorities.

In short: The study tells us that while "doubt" is a major reason for low vaccination rates among migrants, it's not the only reason. For some, it's a lack of trust; for others, it's a lack of access. To solve the puzzle, we need to listen to the specific stories of different groups and offer the right kind of help to each.

Technical Summary

1. Problem Statement

Despite the availability of free COVID-19 vaccines in Germany, significant disparities in vaccination uptake persist between migrants and non-migrants. While structural barriers (e.g., language, legal status, discrimination) are known factors, the specific contribution of psychological antecedents to these disparities remains under-researched.

• Gap: Existing literature has not quantified the extent to which psychological factors explain the gap in vaccination rates between migrants and non-migrants in Germany.
• Objective: To investigate whether migration-related disparities in COVID-19 vaccination are associated with the psychological factors defined by the 5C model (Confidence, Complacency, Constraints, Calculation, Collective Responsibility) and to determine the magnitude of this association.

2. Methodology

Data Source

• Study: "Corona Monitoring Nationwide – Wave 2" (RKI-SOEP-2 study), embedded within the German Socio-Economic Panel (SOEP).
• Period: November 2021 to March 2022.
• Sample: 10,288 respondents aged 18+ (after exclusions) from a weighted sample of 11,162 total participants.
• Design: Cross-sectional survey using multilingual questionnaires (German, Arabic, Farsi, English, Polish, Bulgarian, Romanian) to ensure inclusion of non-German speakers.

Variables

• Outcome: COVID-19 vaccination status (binary: received at least one dose prior to survey start).
• Exposure: Migration status categorized as:
  • Non-migrants (Reference).
  • First-generation migrants (personal migration experience).
  • Second-generation migrants (at least one parent born abroad).
  • Sub-grouping by Country of Origin: Middle East/North Africa (MENA), Eastern Europe, and Other parts of the world.
• Mediators (The 5C Model): Measured via single-item Likert scales (1–5):
  1. Confidence: Trust in vaccine safety/effectiveness.
  2. Complacency: Perception that the disease is not a threat/vaccination is unnecessary.
  3. Constraints: Structural/psychological barriers (hassle, cost, access).
  4. Calculation: Extent of information seeking and risk/benefit weighing.
  5. Collective Responsibility: Willingness to protect others.
• Covariates: Age, gender, education, income, residential area, federal state, prior SARS-CoV-2 infection, and chronic diseases.

Statistical Analysis

• Descriptive Statistics: Weighted proportions and means with 95% Confidence Intervals (CIs).
• Logistic Regression: Multivariable binary logistic regression to estimate Odds Ratios (OR) for vaccination by migration status, adjusting for covariates and adding 5C factors sequentially.
• Decomposition Analysis: The Karlson-Holm-Breen (KHB) method was used to decompose the total effect of migration status into direct and indirect effects. This quantified the percentage of the disparity explained by the 5C psychological factors.
• Software: Stata 17.0 with survey weights to correct for non-response and align with German population statistics.

3. Key Results

Vaccination Disparities

• Overall: 93.2% of the total sample was vaccinated.
• By Migration Status:
  • Non-migrants: 94.3%
  • Second-generation migrants: 88.4%
  • First-generation migrants: 89.1%
• By Region of Origin (First-generation):
  • MENA region: 80.3% (Lowest)
  • Eastern Europe: 84.9%
  • Other parts of the world: 96.5% (Highest, comparable to non-migrants)

Association with Psychological Factors (5C)

• General Trends: Non-migrants reported higher Confidence and Collective Responsibility than migrants. Migrants reported higher Complacency (viewing vaccination as unnecessary).
• Regression Findings:
  • Confidence and Collective Responsibility were strong positive predictors of vaccination.
  • Complacency and Calculation were negative predictors.
  • When Confidence and Collective Responsibility were added to the regression models, the statistical significance of the migration status variable disappeared, suggesting these factors mediate the relationship.

Decomposition Analysis (KHB Method)

This analysis quantified how much of the vaccination gap is explained by psychological factors:

• First-generation migrants: 5C factors explained 61.2% of the disparity.
• Second-generation migrants: 5C factors explained 64.2% of the disparity.
• Dominant Factor: Confidence in vaccine safety was the single most significant mediator, accounting for ~40% of the disparity in both groups.

Heterogeneity by Country of Origin

The explanatory power of the 5C model varied significantly by region:

• Eastern Europe: Psychological factors explained 73.5% of the disparity. The gap was almost entirely mediated by psychological attitudes (low confidence, high complacency).
• MENA Region: Psychological factors explained only 19.4% of the disparity. The direct effect of migration status remained significant even after adjusting for 5C factors, suggesting that structural barriers (e.g., legal status fears, lack of access, discrimination) are the primary drivers for this specific group, rather than psychological hesitancy.

4. Key Contributions

1. Quantification of Mediation: This is the first study to use the KHB method to decompose migration-related vaccination disparities in Germany, demonstrating that psychological factors account for nearly two-thirds of the gap for general migrant populations.
2. Differentiation of Drivers: The study highlights that "migrants" are not a monolith. While psychological factors drive disparities for Eastern Europeans, structural barriers likely dominate for the MENA group.
3. Identification of Key Levers: Confidence (trust in safety) and Collective Responsibility were identified as the most critical psychological levers for increasing uptake.
4. Methodological Rigor: The use of a large, nationally representative, multilingual dataset (RKI-SOEP-2) allows for robust generalizability compared to smaller, convenience-based studies.

5. Significance and Implications

• Policy Intervention: Interventions must be tailored.
  • For groups where psychological factors dominate (e.g., Eastern Europe), communication strategies focusing on building trust, correcting misinformation, and emphasizing collective responsibility are likely effective.
  • For groups where structural factors dominate (e.g., MENA/refugees), communication alone is insufficient. Interventions must address structural barriers, such as legal fears, language access, and ensuring healthcare providers are aware of vaccination rights for asylum seekers.
• Health Equity: Addressing these disparities is crucial not only for pandemic control but for preventing future inequities in routine immunization (e.g., influenza, measles).
• Future Research: The study calls for longitudinal research to establish causality and deeper qualitative investigations into the specific structural barriers faced by the MENA community, which were not fully captured by the psychological 5C model.

Limitations Noted by Authors

• Cross-sectional Design: Cannot establish temporal ordering or causality (e.g., vaccination status may influence psychological attitudes rather than vice versa).
• Selection Bias: Response rates were lower among migrants; vaccinated individuals may be overrepresented.
• Measurement: The 5C factors were measured using single items rather than multi-item scales, potentially limiting psychometric robustness.