Misperceived Discrimination

class: center, middle, inverse, title-slide

.title[
# Misperceived Discrimination
]
.subtitle[
## Opening up a New Field of Research
]
.author[
### Merlin Schaeffer
]
.date[
### 2025-09-29
]

---

layout: true
# Motivation
---

<img src="./media/Gender.png" width="85%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: [Gallup](https://news.gallup.com/poll/1687/race-relations.aspx)]]

???
- Second dimension of Contested Discrimination: Competing views `\(rightarrow\)` competing perceived prevalences.
- Example (not directly related) representing my curiosity:
  + Black and White people agree relations have considerably deteriorated in the US. 
  + But were things really better in 2001?

---

<img src="./media/Anti-Diskriminierungsstelle.png" width="50%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: [Federal Anti-Discrimination Agency](https://www.antidiskriminierungsstelle.de/SharedDocs/downloads/DE/publikationen/Jahresberichte/2024.pdf?__blob=publicationFile&v=2)]]

---

.left-column[
<img src="./media/Anti-Diskriminierungsstelle.png" width="97%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: [Federal Anti-Discrimination Agency](https://www.antidiskriminierungsstelle.de/SharedDocs/downloads/DE/publikationen/Jahresberichte/2024.pdf?__blob=publicationFile&v=2)]]

]

.right-column[
<img src="./media/DW.png" width="100%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: [Deutsche Welle](https://www.antidiskriminierungsstelle.de/SharedDocs/downloads/DE/publikationen/Jahresberichte/2024.pdf?__blob=publicationFile&v=2)]]
]

---

.push-left[
<img src="./media/Freda-LGBTQ.png" width="80%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: [EU-FRA: LGBTIQ Survey III 2023](https://fra.europa.eu/en/publications-and-resources/data-and-maps/2024/eu-lgbtiq-survey-iii)]]
]

.push-right[
<img src="./media/Freda-2024.png" width="80%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: [EU-FRA: Gender-based violence survey 2024](https://fra.europa.eu/sites/default/files/fra_uploads/eu-gender_based_violence_survey_key_results.pdf)]]
]

---

.push-left[
> "The hatred that men bear to privilege increases in proportion as privileges become fewer and less considerable, so that democratic .alert[passions] would seem to burn most fiercely just when they have .alert[least fuel]."

> -- de Tocqueville (1840)

> "[...] greater familiarity with the culture and language and some economic advancement can lead to greater consciousness of .alert[the reality] of discrimination."

> -- Portes, Parker, and Cobas (1980)  
]

.push-right[

]

.push-right[
<img src="https://upload.wikimedia.org/wikipedia/commons/a/aa/Alexis_de_tocqueville.jpg" width="35%" style="display: block; margin: auto;" />
<img src="https://uploads.concordia.net/2025/02/26173752/alejandro-portes-220x220.jpg" width="35%" style="display: block; margin: auto;" />

]
---
layout: false
class: clear

.content-box-blue[.center[
**How accurately do majority and minority citizens <br> perceive ethno-racial discrimination?**
]]

.left-column[
<br>
<br>

.center[We cannot observe <br> *hidden* intentions.]

<img src="https://www.skolske.sk/storage/article/780x520_1588025770_article.jpg" width="100%" style="display: block; margin: auto;" />
]

.right-column[
.push-left[
<br>
<img src="media/Black_Female_Caller2.png" width="35%" style="display: block; margin: auto;" />
]
.push-right[
<br>
<img src="media/White_Female_Caller.png" width="35%" style="display: block; margin: auto;" />
]
<img src="./media/Figure_1.png" width="100%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: Schaeffer, Romarri, Rosenberg, and Krakowski (2025)]]

.center[.font90[
`\(\rightarrow\)` **Limitation**: .alert[Fictitious cases] identify discrimination in the .alert[aggregate]!
]]]

???
4 student confederates called 262 unique municipalities: 860 completed calls

---
class: inverse middle center
# Misperceived *Prevalence* of Discrimination

<br>
<br>
.content-box-green[
**Claim 1**: Mainstream majority members tend to *over*perceive prevalence of minority discrimination.
]

<br>
.left-column[
<img src="./media/Carlsbergfondet_logo_2-liner_UK_RGB_BLACK.svg" width="70%" style="display: block; margin: auto;" />
]
.right-column[.font90[.left[
Schaeffer, Merlin, Krzysztof Krakowski, and Asmus Olsen (2024): “Correcting Misperceptions about
Ethno‑Racial Dis‑crimination: The Limits of Evidence‑Based Awareness Raising to Promote Support for
Equal‑Treatment Policies.” *American Journal of Political Science*: 10.1111/ajps.12933
]]]

???
- First method: Misperceived prevalence.
- Focus on majority members: Native-born Danes with native-born parents.

---
# Misperceived *Prevalence*

.left-column[
<img src="./media/Design.png" width="100%" style="display: block; margin: auto;" />
]

.right-column[
.font80[
Now we are interested in hearing your thoughts on a recent social science study.

**Researchers from the University of Copenhagen conducted a study on discrimination against non-Western minorities in Danish primary schools.** They did this by sending fictitious applications to primary schools in Denmark, where a parent requested to transfer their child to the school's 3rd grade.

**The fictitious applications were exactly the same except for one thing: the name of the father of the child who was to be admitted to the school and who sent the email.** Half of the applications had a typically Danish-sounding name such as "Peter Nielsen," while the other half had a typically Muslim-sounding name such as "Mohammad Osman."

The researchers wanted to find out whether schools would view the exact same application more positively if it was sent by a parent with a Danish-sounding name compared to a parent with a Muslim-sounding name.

**Out of 100 fictitious applications with Danish-sounding names, 25 of the applications were able to get the child admitted to the school.**

**What do you think: out of 100 fictitious applications with Muslim-sounding names, how many of the applications were able to get the child admitted to the school?**
]]

???
- Inspired by two economists: We explain field experiment and ask people what outcome they suspect.

---
# Misperceived *Prevalence*

.right-column[
.center[
<img src="./media/all_misp-2.png" width="85%" style="display: block; margin: auto;" />
]]

???
- We did this with all filed experiments conducted in DK so far: Peter, Asmus, etc.
- Let me help you interpret this.

--
.left-column[
<img src="./media/PredictPrev.png" width="100%" style="display: block; margin: auto;" />
]

???
- In contrast to 3 mechanisms: No evidence of vested interest in perceiving little discrimination.
  + Even among those who are against immigration!
  + Only education: We academics are more off!
---
# Misperceived *Prevalence*

.push-left[
<img src="./media/HaalandRoth.png" width="100%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: Haaland and Roth (2023)]]
]

???
- Same in the US for discrimination against Black people.

.push-left[
<img src="./media/Schaerer.png" width="70%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: Schaerer, du
Plessis, Nguyen, van
Aert, Tiokhin, Lakens, Giulia Clemente, Pfeiffer, Dreber, Johannesson, Clark, and Luis Uhlmann (2023)]]
]

???
- Same for how prevalent we believe gender discrimination to be.

---
class: inverse middle center

# Mis*expected* *Personal* Discrimination

<br>
<br>
.content-box-green[
**Claim 2**: Immigrant minority members rarely expect to be discriminated;<br> but when they do, their expectations tend to be inaccurate.

**Claim 3**: Immigrant minority members tend to over-generalize from prior experiences of discrimination;<br> this is costly.
]

<br>
.left-column[
<img src="./media/dfg_logo_schriftzug_blau_foerderung_en.gif" width="100%" style="display: block; margin: auto;" />
]
.right-column[.font90[.left[
Collaborators: Judith Kas, Pia Hagedorn, and Nina Magdeburg.
]]]

???
- I initially started thinking about this 10 years ago.
- My original question was another one: Are there negative consequences of perceiving too much discrimination?
  + Does it make you unhappy beyond actual reason for it?
- Personal discrimination: **Do _you_ misperceive the discrimination that _you_ experience?**
- Much more difficult: Actual: **Repeated experimental observations** for **real people who can have perceptions**!

---
class: clear
# Central Idea .font70[Mutual evaluation in behavioral games]

---
# Study Design and Sample

.center[
Pre-game survey (n=2,825) `\(\rightarrow\)` Trust games (n=2,010) `\(\rightarrow\)` Pay-out experiment (n=1,880)
]

.push-left[
- Stratified random sample from population registers

- 50% native-born with native born parents,

- Invitation via postal mail.

- Unconditional €20 after survey.<br> Up to €120 overall.

- Online surveys/games in German.<br>Jun '21 to Mar '22.

- Response rate approx. 12%-4%.
]

.push-right[
<img src="./media/Sample.png" width="100%" style="display: block; margin: auto;" />
.center[.backgrnote[*Source*: Schaeffer, Kas, and Hagedorn (2023a)]]
]
---
# The Trust Game

.left-column[
<table class="table" style="width: auto !important; margin-left: auto; margin-right: auto;">
 <thead>
  <tr>
   <th style="text-align:left;"> Sample </th>
   <th style="text-align:right;"> N </th>
   <th style="text-align:right;"> % </th>
  </tr>
 </thead>
<tbody>
  <tr>
   <td style="text-align:left;"> Native mainstream </td>
   <td style="text-align:right;"> 1348 </td>
   <td style="text-align:right;"> 47.7 </td>
  </tr>
  <tr>
   <td style="text-align:left;"> Immigrant </td>
   <td style="text-align:right;"> 888 </td>
   <td style="text-align:right;"> 31.4 </td>
  </tr>
  <tr>
   <td style="text-align:left;"> Desc. of immigrant </td>
   <td style="text-align:right;"> 589 </td>
   <td style="text-align:right;"> 20.8 </td>
  </tr>
</tbody>
</table>

- €10 Endowment `\(\times 3\)`<br> for every €2 entrusted

<br>
.alert[.center[.font120[True first name & city <br> from public registers]]]<br>
.backgrnote[Alisina, Ahmad, Anne, Binyamin, Joyce, Somaia, Sarah, Hayriye, Saibe, Björn, Salem, Fabienne, Sadet, Linda, Margarita, Ali, Joseph, Mhd Kheir, Baran, Bahaa, Jebran, Reno, Seiji, Irina, Ajsel, Christine, Rahim, Yaw Abrefa, Mark, Anjali Dev ...]
]

.right-column[
<br>
<br>
<br>
<br>
<img src="./media/Sender-en.png" width="100%" style="display: block; margin: auto;" />
]

???
- Uni of observation is an interaction, not a person!

---
# LLM-Based Coding of Names
.push-left[
<img src="./media/HuggingFace_2.png" width="100%" style="display: block; margin: auto;" />

.font80[

``` r
library(text)

## 3. Predict ethnicity of first names.
##############################################
zeroshot_ethnic <- text::textZeroShot(
  # The list of first names to be predicted.
  sequences = vornamen,
  # The possible origins for the first names.
  candidate_labels = c("deutsch", "türkisch",
                       "arabisch oder muslimisch", 
                       "ostasiatisch oder chinesisch",
                       "osteuropäisch oder slavisch"),
  # Predict multiple origins for each first name.
  multi_label = TRUE,
  # The template for the hypothesis that is tested.
  hypothesis_template = "Dieser Vorname klingt {}.",
  # The model that is used to predict the gender.
  model = "Sahajtomar/German_Zeroshot")
```
]]

.push-right[
<img src="London25_files/figure-html/unnamed-chunk-26-1.png" width="100%" style="display: block; margin: auto;" />
]
  
---
class: clear
 # 10 Trust Games .font70[Per participant with random partners (2 paid out)]

.push-left[
- *Native mainstream participants*: 
  + 3 games with native mainstream players.
  + 7 games with immigrant origin players.

- *Immigrant origin participants*: 
  + .red[7 games with native mainstream players.]
  + 3 games with immigrant origin players.

<table class="table" style="width: auto !important; margin-left: auto; margin-right: auto;">
 <thead>
  <tr>
   <th style="text-align:left;"> Name </th>
   <th style="text-align:right;"> N </th>
   <th style="text-align:right;"> % </th>
  </tr>
 </thead>
<tbody>
  <tr>
   <td style="text-align:left;"> East-European
sounding name </td>
   <td style="text-align:right;"> 1828 </td>
   <td style="text-align:right;"> 27.2 </td>
  </tr>
  <tr>
   <td style="text-align:left;"> German
sounding name </td>
   <td style="text-align:right;"> 1716 </td>
   <td style="text-align:right;"> 25.6 </td>
  </tr>
  <tr>
   <td style="text-align:left;"> Muslim
sounding name </td>
   <td style="text-align:right;"> 1471 </td>
   <td style="text-align:right;"> 21.9 </td>
  </tr>
  <tr>
   <td style="text-align:left;"> East-Asian
sounding name </td>
   <td style="text-align:right;"> 539 </td>
   <td style="text-align:right;"> 8.0 </td>
  </tr>
  <tr>
   <td style="text-align:left;"> Turkish
sounding name </td>
   <td style="text-align:right;"> 301 </td>
   <td style="text-align:right;"> 4.5 </td>
  </tr>
  <tr>
   <td style="text-align:left;"> Other </td>
   <td style="text-align:right;"> 854 </td>
   <td style="text-align:right;"> 12.7 </td>
  </tr>
</tbody>
</table>

]

???
This setup was important to imply that the composition of partners was the smae in the other group!

.push-right[
<img src="London25_files/figure-html/unnamed-chunk-29-1.png" width="100%" style="display: block; margin: auto;" />
]

???
What did their game partner's entrust them?
- Now focus on games with mainstream partners.

---
# Actual Trust Discrimination .font60[By game partner]

.left-column[
`\(\text{Act Discr}_{ij} = \bar{€}_{j\text{(Germ. name)}} - €_{ij}\)`
]

.right-column[
<img src="London25_files/figure-html/unnamed-chunk-30-1.png" width="100%" style="display: block; margin: auto;" />
]

???
To measure actual discrimination, our identification strategy is a within-partner comparison

The formula shows that for your partner Christina, we take the average she sent to all her German-named partners and subtract what she sent to you.
---
class: clear
# After the 10 Games .font60[Expectations to be discriminated]

.left-column[
`\(\text{Exp Discr}_{ij} = \hat{\bar{€}_{j\text{(Germ. name)}}} - \hat{€_{ij}}\)`

€3 for correct guess of both<br> `\(\times\)` last 7 games

Remember: Composition of partners was 70% outgroup for both native mainstream & immigrant minorities

]

.right-column[
<img src="media/Perception-en.png" width="100%" style="display: block; margin: auto;" />
]
---
class: clear
# Actual vs. Expected Trust Discrimination

---
class: clear
# Actual vs. Expected Trust Discrimination

???
In 67% games immigrants and their descendants did not expect discrimination; <br> 78% they were correct

If they expected unequal treatment, 95% they were inaccurate

---
class: clear
# Actual vs. Expected Trust Discrimination

---
class: clear
.left-column[
<br>
.font150[**Predictors of Misexpected <br> Trust Discrimination**]

.backgrnote[
Bayesian linear model with random effect for repeated games and controls for same city & LLM-based gender, age, & class of name.

Filled estimates: bivariate

Hollow dots: multiple
]]

.right-column[
<img src="media/Report_1.png" width="100%" style="display: block; margin: auto;" />
]
---
class: clear
# The Cost of Misexpected Trust Discrimination

.push-right[
<img src="London25_files/figure-html/unnamed-chunk-33-1.png" width="100%" style="display: block; margin: auto;" />
.center[.backgrnote[Based Selten (1967)'s strategy method.]]
]

.push-left[
<img src="London25_files/figure-html/unnamed-chunk-34-1.png" width="100%" style="display: block; margin: auto;" />

.center[
`\((-0.766€  + -0.401€)  \times 3 \times \frac{2}{3} \times 2 = -4.7€\)`

`\(\frac{-4.7€}{20€} = \color{red}{-23.5\%}\)`

]]

---
class: inverse middle center

# Mis*perceived Personal* Discrimination

.push-left[
<br>
<br>
<img src="https://www.minnisandsmallets.com/wp-content/uploads/2023/03/AdobeStock_145971794-scaled.jpg" width="80%" style="display: block; margin: auto;" />
]

.push-right[
<br>
<br>
.content-box-green[
**Claim 4**: Immigrant minority members tend to<br> over-interpret ambiguous signals as evidence of discrimination.

**Claim 5**: This seems to happen particularly among those with prior experiences of discrimination.
]]

---
# Follow-up Experiment .font60[Disadvantage cue]

.left-column[
n = 1,880

Focus on 878 immigrant origin respondents.

Before full information on overall payout.

3 random conditions: .red[Disadvantage, equal treatment], and advantage.

No deception.
]

.right-column[.font90[
**On average, your ten game partners sent and thereby entrusted 6 € to you.**

In addition, your ten game partners played trust games with other 68 more participants. 25 of these other participants had names that sound typically German.

Below you see a selection of three of these participants with names that sound typically German.

**On average, your game partners sent and thereby entrusted 8 € to these three persons:**

Annegret from Hamburg

Klaus from Berlin

Hartmut from München

**The three have thus received 2 € more than you. How do you rate this result?**

<select class='webex-select'><option value='blank'></option><option value='answer'>My name played a role: I was disadvantaged by my game partners.</option><option value=''>My name did not play a role: I was simply unlucky</option></select>

]]

---
# Follow-up Experiment .font60[Advantage cue]

.left-column[
n = 1,880

Focus on 878 immigrant origin respondents.

Before full information on overall payout.

3 random conditions: .red[Disadvantage, equal treatment], and advantage.

No deception.
]

.right-column[.font90[
**On average, your ten game partners sent and thereby entrusted 6 € to you.**

In addition, your ten game partners played trust games with other 68 more participants. 25 of these other participants had names that sound typically German.

Below you see a selection of three of these participants with names that sound typically German.

**On average, your game partners sent and thereby entrusted 4 € to these three persons:**

Annegret from Hamburg

Klaus from Berlin

Hartmut from München

**The three have thus received 2 € less than you. How do you rate this result?**

<select class='webex-select'><option value='blank'></option><option value='answer'>My name played a role: I was advantaged by my game partners.</option><option value=''>My name did not play a role: I was simply lucky.</option></select>
]]

---
# Follow-up Experiment .font60[Equality cue]

.left-column[
n = 1,880

Focus on 878 immigrant origin respondents.

Before full information on overall payout.

3 random conditions: .red[Disadvantage, equal treatment], and advantage.

No deception.
]

.right-column[.font90[
**On average, your ten game partners sent and thereby entrusted 6 € to you.**

In addition, your ten game partners played trust games with other 68 more participants. 25 of these other participants had names that sound typically German.

Below you see a selection of three of these participants with names that sound typically German.

**On average, your game partners sent and thereby entrusted 6 € to these three persons:**

Annegret from Hamburg

Klaus from Berlin

Hartmut from München

**The three have thus received the same amount as. How do you rate this result?**

<select class='webex-select'><option value='blank'></option><option value='answer'>My name played a role: In this case I nevertheless received the same.</option><option value=''>My name did not play a role: I was treated equally by my game partners.</option></select>
]]

---
class: clear
# Perception of Discrimination .font70[My name played a role ...]

---
class:  clear

.left-column[
<br>
.font150[**Effect of Disadvantage Signal Compared to Baseline**]
]

.right-column[
<br>

<img src="media/Report_3.png" width="100%" style="display: block; margin: auto;" />
]

---
layout: false
# Conclusion .font70[for now ...]

.push-left[
<br>

- Because discrimination is often concealed, people struggle to accurately generalize their experiences to new situations and over-interpret signals of disadvantage.

- This, in itself, results in additional downstream mental and economic costs.

<br>
.font140[.center[
.alert[Thank you for your attention!]
]]]

.push-right[.content-box-green[

**Claim 1**: Mainstream majority members tend to *over*perceive prevalence of minority discrimination.

**Claim 2**: Immigrant minority members rarely expect to be discriminated;<br> but when they do, their expectations tend to be inaccurate.

**Claim 3**: Immigrant minority members tend to over-generalize from prior experiences of discrimination;<br> this is costly.

**Claim 4**: Immigrant minority members tend to<br> over-interpret ambiguous signals as evidence of discrimination.

**Claim 5**: This seems to happen particularly among those with prior experiences of discrimination.
]]

---
layout: false
count: false
# References

.font70[
Haaland, I. and C. Roth (2023). "Beliefs About Racial Discrimination and Support for Pro Black
Policies". In: _Review of Economics and Statistics_, pp. 40-53.

Portes, A., R. N. Parker, and J. A. Cobas (1980). "Assimilation or Consciousness". In: _Social
Forces_, pp. 200-224.

Schaeffer, M., J. Kas, and P. Hagedorn (2023a). "The Association between Actual and Perceived
Discrimination (APAD): Technical Report". In: _SocArXiv_.

Schaerer, M., C. du Plessis, M. H. B. Nguyen, et al. (2023). "On the trajectory of discrimination: A
meta-analysis and forecasting survey capturing 44 years of field experiments on gender and hiring
decisions". In: _Organizational Behavior and Human Decision Processes_, p. 104280.

Selten, R. (1967). "Die Strategiemethode zur Erforschung des eingeschränkt rationalen Verhaltens im
Rahmen eines Oligopolexperiments". In: _Beiträge zur experimentellen Wirtschaftsforschung_. Ed. by H.
Sauermann. Tübingen: Mohr, pp. 136-168.
]