Can a Shared National Identity Trump Ethnic Identity? Evidence from a Field Experiment in Singapore

Last registered on July 08, 2021

View Trial History

Pre-Trial

Trial Information

General Information

Title

Can a Shared National Identity Trump Ethnic Identity? Evidence from a Field Experiment in Singapore

RCT ID

AEARCTR-0007559

Initial registration date

July 08, 2021

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

July 08, 2021, 3:00 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Singapore

Region

Primary Investigator

Name

Guillem RIAMBAU

Affiliation

Universitat de Barcelona (UB) i Institut d'Economia de Barcelona (IEB)

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Risa Toha

PI Affiliation

Yale-NUS College

Contact Investigator

Additional Trial Information

Status

In development

Start date

2021-07-01

End date

2022-12-31

Keywords

Behavior

Additional Keywords

Pro-social behavior; altruism; discrimination; out-group vs. in-group; social identity.

JEL code(s)

D91; J15.

Secondary IDs

Abstract

Ethnic diversity has been widely associated with lower levels of social harmony. The main reason for this
is the proven existence of ethnic biases, which have been shown to impact romantic preferences, neighborhood choices, job recruitments, electoral votes, and political stability, among others. Oftentimes, however, the ethnic dimension cannot be easily disentangled from other dimensions of identity, such as class or nationality. Hence, the independent effect of ethnic diversity remains to be fully understood. This project aims to fill in this gap: by leveraging Singapore’s unique demographic composition, it aims to pin down the independent effect of ethnic, national and socioeconomic identities on social harmony. To be precise, the goal is to measure the extent to which they impact (if at all) pro-social, altruistic behavior. To do so, we conduct a field experiment, using the “misplaced envelope technique”. This consists of sending envelopes with valuables inside (in our case, a voucher) to the wrong addressee. The actual recipient, upon realizing that (s)he is not the intended recipient, can choose to keep the voucher or return it. By randomizing the ethnicity, nationality and socio-economic status of our actual and intended recipients, we can assess which dimensions trigger altruistic behavior (i.e., returning the voucher to the correct addressee) over others. Our study bears important contributions. Theoretically, uncovering how these different affiliations affect altruism will help us understand broader questions about the process through which national integration and social cohesion can be cultivated in a multiethnic context. Methodologically, to the best of our knowledge, our study will be the first to analyze this issue by using observed behavior, as opposed to survey data, hence overcoming social desirability biases that arise in surveys. In terms of policy, our findings should shed light on how social integration and cooperation across groups can be strengthened.

External Link(s)

Registration Citation

Citation

RIAMBAU, Guillem and Risa Toha. 2021. "Can a Shared National Identity Trump Ethnic Identity? Evidence from a Field Experiment in Singapore." AEA RCT Registry. July 08. https://doi.org/10.1257/rct.7559-1.0

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

The goal of this study is, by leverage Singapore’s unique demographic composition, to pin down the independent effect of ethnic, national and socioeconomic identities on pro-social, altruistic behavior. To do so, we conduct a field experiment, using the “misplaced envelope technique”. This consists of sending envelopes with valuables inside (in our case, a voucher) to the wrong addressee. The actual recipient, upon realizing that (s)he is not the intended recipient, can choose to keep the voucher or return it. By randomizing the ethnicity, nationality and socio-economic status of our actual and intended recipients, we can assess which dimensions trigger altruistic behavior (i.e., returning the voucher to the correct addressee) over others.

Intervention Start Date

2021-07-01

Intervention End Date

2021-12-31

Primary Outcomes

Primary Outcomes (end points)

Willingness to return the voucher

Primary Outcomes (explanation)

Willingness to return the voucher is measured as a categorical variable: 1=intends to return the voucher (contacts sender using information in the misaddressed letter); 0=uses the voucher himself/herself (we can check this thanks to an agreement with the issuing company); 99= There is neither contact with the researchers nor usage of the voucher (so presumably the voucher was forgotten, discarded as junk mail, or thrown away).

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

The experiment is based on the misaddressed letter technique. Residents will receive an envelope addressed to them. However, the letter (and token of appreciation) inside will be addressed to someone else (seemingly by mistake). We will examine which resident and addresse characteristics explain willingness to return the voucher to the (perceived as) correct addressee.

Experimental Design Details

A survey on social mobility and perceptions on Singapore will be carried out on a representative sample of 3150 local residents. Unbeknownst to them, residents will receive a letter a few weeks after the survey. The name on the outside of the letter will be the correct one (we will gather their names during the survey collection). However, the letter inside the envelope will be addressed to someone with a different name. The envelope will also include a voucher for the local supermarket Sheng Siong (either $10 or $50) which is supposed to be a token of appreciation for the recipient. Since the letter will contain the sender's contact information, we can track when residents intend to return the voucher because they realise it was not meant for them (they can contact the sender by phone, email, or return the sender's address). Importantly, we can distinguish those respondents who keep the voucher because they want to use it from those who keep the voucher unknowingly because they don't open the letter: there is an agreement with Sheng Siong to let us know where and when all such vouchers are used.

Randomization Method

The raffle will be allocated via randomization done in office by a colleague computer.
Randomization of misaddressed letters will be done via die roll taking into account that some quotas have to be met.

Randomization Unit

The units of randomization will be individuals.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

(design not clustered).

Sample size: planned number of observations

3150 individuals receiving a letter

Sample size (or number of clusters) by treatment arms

3150 individuals receiving a letter

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

[All computations using power twoproportions average1 average2, n(#) onesided in STATA16] For our recipients getting $50 vouchers, comparison groups will be of size 150. This means that for an effect size of 10%, for one-sided comparisons, we have a statistical power ranging from 0.5384 (if return averages are 0.5 & 0.6) to 0.9901 (if return averages are 0 & 0.1, OR 0.9 & 1). Power reaches nearly 80% at the 0.8 vs 0.9 return rates (0.7847). For our recipients getting $10 vouchers, comparison groups will be of size 300. This means that for an effect size of 10%, for one-sided comparisons, we have a statistical power ranging from 0.69 (if return averages are 0.5 & 0.6, or 0.4 & 0.5) to 0.9999 (if return averages are 0 & 0.1, OR 0.9 & 1). Power reaches surpasses 80% at the 0.7 vs 0.8 OR 0.2 vs. 0.3 return rates (0.81). Last, for our recipients getting $10 vouchers, comparison groups will be of size 300. This means that for an effect size of 5%, for one-sided comparisons, we have a statistical power ranging from 0.33 (if return averages are 0.5 & 0.55, or 0.45 & 0.5) to 0.99 (if return averages are 0 & 0.05, OR 0.95 & 1). Power reaches nearly 80% at the 0.9 vs 0.95 return rates (0.75).

Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

IRB